RotoHog Football
Sortable RotoHog Stats
Weekly RHP

Basketball
Hoop Pointers
Player Lookup

NBA Schedule
Full Season
Next 4 weeks
Sched-O-Matic

TSN Hoops
TSN Stats
Yesterday's TSNP
Team Rankings
Assimilator

NBASE
NBASE Stats
Yesterday's RHP

Market Madness
Contest Site
Rules
Historical Stats
Recap of Entries
Unit Returns
Standings

Baseball
MLB Schedule
Base Advances
Player Lookup

TSN Baseball
Sortable Stats
Yesterday's TSNP
Team Rankings
Assimilator

RotoHog Baseball
Sortable Stats
Yesterday's RHP

Draft Bug Baseball
Sortable Stats
Yesterday's DBP

General
Home page
Hall of Fame
Register
Gur-who?

Each team plays 82 games. Over 170 days, that means a player (who avoids injury) will appear in games slightly less than 50% of all paydays. (82/170=48.2%)

You are required to hold a player for at least 3 days. Over any 3 day period, roughly half of all teams (on average) play twice. So, one strategy would be to pick up players who play 2 out the next 3, and drop players who play less than that. Is that efficient?

Let's do a cost-benefit analysis. Assume for this example that you're always considering among players with equivalent production per unit cost. (Clearly, that won't be true, but otherwise the analysis gets clumsy.) Option A is to hold John Doe throughout the season, during which time he has 82 gamedays and costs 170 paydays, producing an "efficiency rating" of 48.2%. Option B is to actively drop a player whenever he has an off-day and he also plays only once in the next 3 days. I think this would require at most 55 trades, so you'd spend no more than an extra 55 paydays, or 225 paydays in total. But you would get 112 gamedays for that cost, for an efficiency rating of at least 49.8%, which is slightly better. While you would have to own players with lower daily costs (since you'll have to pay for 225 days instead of only 170 days), as long as the pts/game/$ are equivalent, it's worthwhile to do the trades.

In real life, you should be able to maneuver through a better player "rotation" pattern than the simplistic 2-for-3 swap every three days. For example, if you could generate a regular cycle of 3 games every 4 days, the efficiency rating improves to at least 60.3% (128 games/212 paydays). Maintaining a 3/4 game/day ratio may be a bit aggressive - a quick scan of the schedule shows that over all consecutive 4-day periods, an average of only 4 teams have 3 games scheduled - but it might be a good benchmark to aim for. And if consistently executed, it produces a 25% increase in expected points over the season (vs. a no-trade strategy).

If you're going to aim for a 25% productivity increase by actively trading, then you should probably not trade players who are producing points 25% better than their locked salary would suggest. I'm sure there will be some players who appreciate much more than 25% in the early weeks, and that's why your first priority should be to find them quickly. A viable draft strategy might be to fill your draft roster with players who you think are the most undervalued. After a few games, or perhaps a few weeks, you can prune out the ones who didn't materialize as expected, and use those slots for active trading.

Wow, this turned out to be a lot longer than expected! Hope it's helpful.

Yesterday was another maddening day of technical difficulties, as I lost CGI capabilities on yet another server. The outage occurred shortly after new SW football prices were posted, and I suspect I may have blown a "circuit breaker" of some sort, but I have yet to get any feedback from that host. Hopefully I can get to the bottom of it. That server was hosting the football stats and the Football Pickoff game. I had the stats reports successfully transferred to another server within a few hours, and Football Pickoff was successfully relocated late last night. The new links are properly coded on this page now. I hoping this new server will be more reliable (RotoGuru2.com, which I've affectionately dubbed "The Deuce"), but with two server knockouts in two weeks, I'm understandably nervous about next week, when daily hoops repricing gets overlaid on top of continued football reporting. If it blows up, Wednesday will be the day. Cross your fingers, because as of right now, I don't have any more contingency fallbacks.

Speaking of football pricing, it looks like many of you have gotten the hang of it, as the median value gain for the RotoGuru ranked teams was almost $5.5 million, and the median franchise value now exceeds $80 million. The trick is now to put all of that wealth to productive use.

On to the Basketball tip of the day. Most of the preseason uncertainty for Smallworld Hoops centers around the impact of daily repricing. Over the last few SW games, we've gotten a good feel for how "the masses" seem to trade with weekly repricing, but now the timing of those trades during the week will have an impact. Here are my advance thoughts:

• New trades will still be doled out weekly, on Thursday afternoons. Evidence from other games suggests that many managers spend their new trades as soon as they are available. I see no reason why daily repricing should change that pattern. This suggests that the Friday and Monday repricings will probably incorporate the impact of the lion's share of trading.

• Remember, however, that higher trading volume does not translate into larger price swings. As we learned during the SW baseball season, the magnitude of price changes for any repricing are proportionate to the total number of transactions included in that cycle. If there are twice as many trades done on Friday vs. Tuesday, then 2 Friday trades will have the same price impact as one Tuesday trade.

• Therefore, there is no reason to expect the greatest price changes will occur on Friday or Monday. If those days do experience the highest activity, then the results will probably be similar to the results we saw in last winter's SW Hoops game, where upcoming schedule intensity seemed to be one of the most influential price determinants. (Players with heavy upcoming schedules showed gains, and those with light schedules were sold.) This may therefore make the Friday-Monday changes more predictable. But it won't necessarily produce larger gains for the top players.

• It is difficult to predict how midweek (Tues-Thurs) trading will behave. Will this continue to reflect schedule anticipation, or will it show more of a "big game" response? Since trading on those days is likely to be much lighter, I expect repricing on those days will be much more reflective of the "theme of the day" - whether that be a schedule event, or a notable individual game, or an injury, etc. If so, these days could actually produce greater price volatility at the extremes, since fewer trades will be needed to produce a maximum price move.

With this outlook, one of the best ways to position your draft roster is to identify those players that are most likely to benefit from early season buying. This could point to players on teams with heavy early schedules. On the other hand, if everyone starts out the season loaded with Orlando players (since the Magic play 7 games in the first 10 days), this will dampen the price upside of those players (remember that preseason draft buys do not push up prices), and will also heighten their price downside once their schedule advantage dissipates. If you overload with these players, you may have troubling getting out of the way when the sell pressure occurs.

A contrarian approach might be to draft some players on lightly scheduled teams. Conventional wisdom would steer most managers away from players on Minnesota or Sacramento, since neither team plays in the first three days. However, both teams have back-to-back games on Friday-Saturday, and if these players get a lot of buying support over that first weekend, they could have some nice early upside. Still, neither team has an attractive schedule for the balance of that trading week, so it is also possible that they will be ignored in favor of other teams that have 4-5 games scheduled over the first full trading week.

I often point out that the best way to learn how to do something well is to really screw it up a few times. I see no reason to expect this season will be any different. Perhaps the best tip is to stay tuned to the daily blurbs and the message forum as the season unfolds, so we can all learn from the collective Gurupie experience, rather than taking our early lumps independently. Sound like a plan?

Today's Basketball tip of the day is an obvious one: Watch the schedule. From week to week, there are wide disparities in the number of games played. One important way to maximize your points is to maximize your exposure to games played. This applies to both covered games, but in different ways.

For SmallWorld, schedule watching is important not only to maximize your game exposures, but also to improve your value generation. Since many others will also be buying players in advance of heavy game schedules, this is an important way to identify likely price gainers. And the longer a player's heavy schedule persists, the more likely that player will be able to hold his value gains for an extended period.

In Echelon's game, you pay your roster of players daily, whether they have a game scheduled or not. So it can be costly to hold on to players who don't have favorable schedules. Granted, there will probably be some players who are cheap enough to be worth holding through thick and thin. But for a good portion of your roster, you will want to actively rotate players to optimize game exposure vs. cost.

I provide several tools which should make schedule planning much more efficient. The color coded schedule makes it much easier to identify teams with relatively light and heavy schedules. Look for the teams with a long period of green and white dates, and avoid those teams with frequent bursts of red. I also offer the NBA Sched-O-Matic which allows you to customize your slicing and dicing of the schedule to match your schedule outlook period. Don't rely on either of these tools exclusively, because they each offer a different perspective.

Here are a few tidbits to get you started:

• Neither Minnesota nor Sacramento play at all for the first three days of the season, and they each have light schedules for awhile afterwards as well. If you've got your eye on Garnett, or Webber, or even Wally Szczerbiak, it probably makes sense to wait on those pickups.
• At the other extreme, Orlando plays 7 games in the first 10 days. Not only will that provide a lot of points per dollar in the early going, but think about how they will appear in Smallworld's early season top 50 lists.
• While Orlando has the strongest start, teams like Houston, New York, the Lakers, Seattle, and San Antonio also have favorable schedules. In fact, those teams may even be marginally more attractive for Smallworld, because although they play one less game (vs. Orlando) in the first 10 days, they have a more attractive schedule for the following week, which should give their players more prolonged price protection.

Today I'll begin a series of Basketball Tips of the Day, as the NBA season is now only one week away. My first tip may seem trivial, but it may also be the most important: "Read the Rules". Both the Smallworld and Echelon games have some significant changes from their prior hoops games. If you are going to fashion a productive draft, you really need an advance understanding of the player characteristics you are targeting.

In Smallworld, we have fewer trades and daily repricing, and with a daily price cap of $1 million, I expect significantly more price volatility, which makes the upside potential greater, but also heightens the downside risk. Our weekly dole of trades is only 4 (last season it was 5), which means that you'll only be able to turn over your roster ever 2.5 weeks. Last year, holding a player for two weeks only required holding him through 2 repricings. This year, a 2-week holding period will span 10 repricings. While we will have 8 trades for the first 4 days (4 to start the season on Tuesday, and another 4 after the Thursday repricing), you'll want to make sure you are efficient in getting the most for your trade usage, and that will require a reasonably thoughtful draft. More on this later in the week.

For Echelon, the format is quite different from recent Echelon offerings (although very similar to their current hockey format). If anything, the changes make planning even more paramount. You start the season with a fixed pot of cash from which you have to "pay" your roster each day over the course of the year. If you run out of funds too soon, your season ends prematurely. And if you have funds left over at the end of the season, it is wasted. While this format allows great flexibility in pacing your season, success will require more patience and planning than is typical in many fantasy sports games. But this should work to the advantage of most Gurupies, who tend to be more thoughtful about their game plans, and who are more comfortable with the technical nuances. In this game, trade capacity is virtually unlimited (each player is subject to a minimum holding period, however), and while that probably makes your initial draft less critical to long term success, identifying and locking in a few bargains early on will be very important.

Even if you read the rules when you first registered, it's a good idea to go back and reread them now. They aren't that long! And it's common for preseason rules to undergo some changes and clarifications after the first release. You may very well notice something that you don't recall from your first reading.

So go do your homework, in preparation for tomorrow's tip, which will be meatier - I promise.

Rich Gannon topped the list of offensive SWP producers, while his favored target Tim Brown headed the list in the Echelon game. Brian Griese had another strong day as the second best QB in both scoring systems. And Green Bay ran away with the defensive honors, topping the next best defense by more than 100 SWP, largely on the strength of 6 interceptions against the Chargers.

In Football Pickoff, if you had decided that the "no-brainer" strategy was simply to pick the consensus underdogs, you got your comeuppance yesterday. The approach netted only 3 winners through 13 games, "good" for -157 points. In fact, the best strategy for this week looks like going with the betting favorites, as only two entries posted better than the "Vegas Favorites" score of +381. With only one game left, 298 entries have positive points for this week, roughly two-thirds of the active field. The top two entries for the week each have 12 correct picks out of 13 games.

A note about YTD scoring in FP is warranted. There were about 50 new entrants who registered during the past week. For YTD scoring, I assessed these entries with the -500 pt. wimp penalty for week 2, even though they didn't start until week 3. However, for all of the sortable scoring averages, I am using only scores since first entry. So, while late entrants are hurt in the overall point standings, they can still get a feel for their relative "in game" success by sorting on YTD average. Clear as mud?

Tomorrow will be my first preseason Hoops Tip of the Day. The NBA season starts a week from tomorrow. Time to start thinking roundball.

You will notice that you can't make any post-game changes to your pick for last night's game in Football Pickoff. You can click on the KC button, but it won't do any good. Odds on half of the games this week remain inside of 60/40, so picking underdogs is going to involve more risk this weekend. Remember, the wimp penalty is still in place, so you need to pick at least 10 games.

I haven't gotten any stats updates this morning. With only two teams involved, if you're curious, work out the points on your own. It looks like Smallworld has updated points this morning, but Echelon doesn't update points until Tuesday morning. A quick scan of the boxscore suggests that Tony Gonzalez was probably the most notable point producer of the game, with 220 SWP and 592 Echelon points. The KC defense also had a very respectable 283 SWP.

I think I have all of the new football stats links switched over to the new server now. If you find a link that still tries to go to the old (Hypermart) server, first try reloading the referring page, just in case you're getting an older cached version. If the old link remains, let me know so I can update it. For more details on the reason for the change, see yesterday's blurb.

I should have the Hoops Assimilator posted sometime this afternoon. I'm adding the ability to "remember" cash balances for the SW and CNN games, although you'll have to manually revise the cash number to accommodate things like ad click bonuses. There will inevitably be a few minor bugs that escape my attention. Please report them as you discover them. I've just taken the football version and reconfigured it for basketball, so there may still be some unwanted football remnants that will need to be adjusted.

Next week, I'll begin producing some preseason "basketball tips of the day", as I did prior to the football season. If you have any thoughts about specific issues that I should address, send me an email.

Since the stats have been moved to a different server, the links have been changed. I have replaced the links on this page, but it may take a few days to get all of the associated links changed on other pages, so please bear with me. If you have separately bookmarked any football stats URLs, you will need to update your bookmarks. I expect to leave the football stats where they currently reside, at least for the time being. I do need to look around for an alternative backup server, however, in case lightning strikes twice, so this may not be the "permanent solution" - if there is such a thing.

With the server issues to attend to this morning, I'm behind schedule on the release of the Basketball Assimilator, but I still hope to have it ready within a day or so. Fortunately, time is not yet of the essence for that application.

SW price changes yesterday were quite concentrated. Nine players posted the maximum gain of $1 million at the SW site, and ten reached that mark at the CNN site. A lot of this was due to the high number of reasonably priced high producers who had just returned from their bye week, so I don't necessarily expect such concentrated repricing in the future. The median value increase for the RotoGuru ranked teams was just above $5 million, and the median franchise value now exceeds $75 million.

As evidence that franchise value is no longer a significant constraint for many teams, the average cash balance for the RotoGuru ranked teams last night was $5 million. Even more telling, over 20% of the teams were lugging in excess of $7.5 million, an amount which exceeds the value of the average player on these rosters. While there will continue to be obvious gains-generating opportunities, if you are already flush with cash, you need to consider whether the player you'd like to pick up for gains is really the player you want to own this week for points. If you can find more attractive matchup opportunities, you may want to forego the extra expected gains in favor of striving for more points. After all, the only reason to seek gains is to be able to afford the roster you want.

Football Pickoff "odds" made some larger than normal shifts yesterday. Several games experienced shifts in the neighborhood of 15%. Two games flipped favorite/underdog status, with Green Bay now favored over early favorite San Diego, and Carolina now preferred over Detroit. The shifts were mostly the effect of new picks being made, rather than old picks being changed. But if you made your picks early, you may want to stop by and see how your relative risk looks now. Remember that picks for tonight's game will be frozen at game time, but the remaining games can be changed until 1pm EST Sunday.

I hope to have the basketball Assimilator outfitted with the new prices and stats by tomorrow. That's my project for today. There are less than 2 weeks to go until the NBA season starts. I guess I'd better start thinking about putting together a roster, rather than just putting together a web site.

Football Pickoff odds are shaping up to make this a difficult week. (...as though the last two weeks weren't difficult enough!) Half of the games are now listed at 60/40 or tighter. Last week, only two games closed in that range. It looks to me like the "market" is getting more efficient.

If you are waiting until after repricing to make your roster moves this week, don't dally. Get it done by tomorrow morning. You can fiddle with the other 13 games of your Pickoff entry until Sunday, but your Smallworld and Echelon rosters will be frozen tomorrow.

SW Football rankings will be updated after the repricing today.

I've been asked whether the Thursday game this week, and the resulting early roster freeze, is likely to impact the trading patterns for this week's SW price change. Thus far, trades between Wednesday and Sunday seem to have dominated the flows. If this is because many teams need to wait until Wednesday to get their next allocation of trades, then the change in schedule should have no effect this week. If late-week trading is due to general procrastination, then there could be an impact. If there is a noticeable impact, it would probably show up as early dumping of week-7 byes, which include Arizona, Jacksonville, and Tennessee. If you have players or defenses from any of those teams, it's worth considering whether you want to wait. My best guess is that there won't be much impact.

In Football Pickoff, five entries this week did better than simply picking the underdog in each game. The top entry for the week, Super Redhawks, had the consensus underdog in every game except for the Washington-Arizona game. Two of the other entries in the top 5 picked 12/14 winners correctly. Of the 14 games this weekend, underdogs won 8. If that trend continues, picking underdogs will be a tough strategy to beat. Of course, most entrants can't pick the consensus underdogs, by definition.

Some people feared that by posting the emerging pick percentages each day, most teams would wait until the last minute to make their picks. However, that was not the case. Roughly 20% of the entries did make or change their picks in the final 12 hours. But none of the selection percentages changed by more than 2% during those final hours, so late picking really didn't have much of an impact. In fact, only one game shifted by more than 10% over the course of the entire week. Early in the week, the 49ers were a 73/27 favorite over Carolina. By the end of the week, the gap had narrowed to 61/39.

And with the Giants coming through last night, I eeked out a positive score of +4, even though I had only 4 games picked correctly. It may be time to start throwing darts myself.

If you've hung onto Marshall Faulk in spite of his previous two weeks, your patience was nicely rewarded. If not, chances are that Faulk outpointed your entire cadre of running backs. And most of the top receivers also had off days. At least Marvin Harrison caught a few passes. Terry Glenn was shut out, and it appears that Yancey Thigpen didn't even play.

And who'd have guessed that the best defense of the day would be the team facing Green Bay?

A strange week. If you think you scored well, then you probably scored VERY well relative to most teams.

In Football Pickoff, about 30% of the entries posted positive scores this week, a threefold increase over last week. If you went strictly with the underdogs, you'd have ranked #10 for the weekend so far. Of course, if more people start picking underdogs, then the risk/reward tradeoff for the underdogs becomes less attractive. I expect the market will become more and more efficient at getting the percentages appropriately calibrated, as this week's results clearly indicate. I'm going to stick with the same ground rules for next week: same scoring, and daily updates of pick percentages.

You will note that I have ranked entries based on YTD average in addition to cumulative points. This gives late starters a better opportunity to gauge their relative performance. However, to be eligible for the automatic Hall of Fame berth, you must have competed in every week. And that info is also displayed.

The NFL schedule provides an unusual Thursday game this coming week. As such, you'll have to get your fantasy rosters set up earlier than usual. Don't get caught napping! For Football Pickoff, you must have the Thursday game selected before game time on Thursday, but you may wait until Sunday for the rest of the games. I'll just lock out your ability to change the Thursday pick after that game starts.

I updated the Pickoff selection percentages late last night. In general, they remained pretty stable compared with the prior day. More than half of all registered game players have now submitted picks. You have until gametime to submit or change your picks, but I would advise against waiting until the last minute. You never know when servers will bog down.

I should make a clarification about the www.hoopsklyce.com site which I mentioned in Wednesday's blurb. I had mentioned that this site offers player projections tailored to Smallworld scoring, and it does,...   but that's just one option. They also offer rankings based on other scoring systems, including Echelon's point formula, traditional rotisserie-type scoring, and will customize a report for any new scoring format you might encounter. The site also offers daily email reports with the latest notes and observations.

If you plan to submit your SW football team for my rankings, please get the info in to me by the end of October. (If you need instructions, see the bottom of the rankings table.) I'm going to stop adding teams once November rolls in. In prior sports, new team submissions kept straggling in throughout the season, but it's a nuisance to keep adding teams, and it's not fair to others to keep introducing new teams, particularly if they are near the top of the rankings. If you're shooting for that coveted Hall-of-Fame berth, you need to know who your competition is before the last few weeks of the season. So if you are one of those chronic procrastinators, get it in soon, or else forget it.

Speaking of the Hall of Fame, I still haven't heard from most of the baseball winners. It's not easy for me to track down who submitted which teams (I can do it, but it is a tedious task), so if you are one of the winners, please send in a draft of your proposed enshrinement blurb.

Updated team rankings were posted yesterday evening. The median increase for the RotoGuru ranked teams was about $4.7m, and the median franchise value now stands just above $70 million. The roster of top producers (see yesterday's blurb), which was valued just below $66m yesterday morning, now costs $68m.

For the past two trading weeks, bye-related trading has dominated the flows, and to some extent that will happen next week, too, as teams will now be reclaiming some of the hot Washington, Seattle, and Carolina players who just got much cheaper. Remember, this is a zero sum game, so for every time that one of these players is added, someone else must be dropped. It's not necessary to have a bad performance to induce a loss; you just need to be less attractive than some other player.

Turning to Hoops, I now have stats tables posted with draft stats and prior regular season stats. The point formulas for both the Echelon and Smallworld games are the same as last year's games, so I was able to include some late season period breakouts as well. Since prices are generally based on performance over the full prior season, those players who got off to slow starts might be relatively more attractive - unless, of course, they get off to a slow start once again.

I've tried to review and correct all of the team affiliations, but this is a manual process, so errors are inevitable. For players not currently signed, I'm showing the most recent team affiliation, rather than showing their free agent status. Please let me know if you find any mistakes.

With the stats and schedules now updated for the new season, the next Hoops project is to update the Assimilator. However, this will not get done until next week.

Today's programming project is to work on adding division capabilities to the Football Pickoff game. Speaking of which, I updated the selection percentages as of 11pm last night. So far, only one team has been picked more than 80% of the time - Jacksonville (93%) over Cleveland (7%). At this juncture, it looks like roughly 40% of the picks are in, and of course even those picks are subject to change.

Keeping the focus on hoops for the moment, I'm close to having my draft stats tables prepared. Hopefully they'll be posted by tomorrow. But they will only show this year's prices vs. last season's stats, and with a lot of new players, some key trades, and aging, some players will obviously do much better that last year, and vice versa - which is why it's more important to assess how players will do this year, rather than how they did last year. I've noticed one site that provides player projections and rankings tailored to the SW game, which is www.hoopsklyce.com. The cost of the projections is nominal, and it could be just the thing to jump start your season... especially if you don't have the time to do all the research yourself.

Turning to football, I worked up the top performing SW football roster, based on SWP/game through the first 5 weeks. Here it is:
QB: Warner, Johnson
RB: Davis, Biakabutuka, Levens
WR: Harrison, Bruce, Westbrook
TE: Walls
KR: Kirby
K: Mare
Def: Jacksonville
Total SWP/G after 5 weeks: 3272
Total cost on 10/12/99: $65,750,000

Remarkably, this roster costs just under $66 million, with only 3 players costing more than $7 million, and none of the running backs priced above $5 million (although that will change this afternoon). Unless some of the big names start to kick it into high gear, the profit motive for trading is going to diminish rapidly, as the extra value isn't going to have much use. Roughly half of the teams in the RotoGuru rankings can already afford this team.

Time to plan your weekly Smallworld Football trading strategy. Since some of you don't seem to pay attention very well, let me say it again: Don't pick up week-5 bye players until after this Wednesday's repricing! If you do plan to trade for some of those players, and you need to do the sell side of the trade before Wednesday (to duck a probable loss), then leave the slot empty through the repricing. There is no penalty for an incomplete roster until Sunday.

You can now make your Football Pickoff selections for week 6. I've added one new scoring rule this week, designed to ensure that you can't succeed too well by avoiding picks. Details are at the game site. I'll also be adding the capability to create and join divisions this week, but that feature won't be ready for a few more days.

I'm going to try to get some preliminary basketball draft data posted in the new few days as well. The NBA season doesn't start until November 2nd, but I'd like to have something available fairly soon to assist you in your preseason mulling.

If you finished in the top spot in any of the RotoGuru baseball team rankings, you qualify for entry into the RotoGuru Hall of Fame. I'd like those people to send me a draft paragraph about themselves, similar in scope to the write-ups that accompany the existing HOF-ers. I reserve the right to edit these as I see fit, but I need you to take the first shot. In addition to the five posted games (SW full and mid, CNN full and mid, & Echelon BP), if the winner of Echelon's Baseball Market game is also a Gurupie, I'll admit him/her as well. Same thing goes for Echelon's Football games. I haven't been posting separate rankings for those games, because most Echelon players are Gurupies, and because Echelon already does a good job of producing their own rankings, so my efforts wouldn't seem to add much value.

Kurt Warner looks like he is this year's answer to last year's Randall Cunningham. He was expected to be a backup QB this season, but got thrust into the starting role because of an injury, and just keeps throwing up huge numbers - for a very cheap fantasy price. Was he ever this successful in arena football?

Football points for yesterday were posted by about 8:30 am today. A quick comparison indicates that my numbers and Smallworld's seem to agree, although that is based on a very limited sample. Let me know if anything looks amiss.

I suspect a lot of Football Pickoff players feel something like the Cleveland Indians after last night's debacle. Hang in there. I've always said that the best way to learn how to do something right is to really screw it up the first time. Just think how much smarter we all are now! I think the scoring carnage is generally a byproduct of two interrelated factors:

1. No one could see how lopsided many of the picks were, which introduced a lot a risk, and
2. Many entrants failed to appreciate the risk/reward tradeoff of picking a favorite.

I redesigned some of the game site yesterday, adding some new features, and also enhancing the look. Still much more to do in that regard, though. I also plan to introduce the ability to join a league this week, so you can see your score against a smaller universe of players. Please be patient as I add new features. I want to make sure I get it right.

By the way, I made a few changes to the NBA Sched-O-Matic so that it should now run properly for web TV users. Thanks to Gurupie Paul Sheirer for pointing me in the right direction.

I will post the actual selection percentages on Sunday afternoon, probably by halftime of the first games. By the way, make sure you get your picks finalized before 1:00 PM EST on Sunday. The system will still record your picks after that time, but your score will only reflect activity prior to the opening kickoff.

I've been advised that the Football Pickoff server will be taken offline for three hours for some hardware upgrades. This is supposed to take place between 4:00-7:00pm EST either today (Friday) or tomorrow (Saturday). Hopefully, there won't be any lingering complications. Plan accordingly.

I was alerted this morning to a few obvious errors in week 17 of my NFL Schedule. These have now been corrected, but if you have a printed version, you probably ought to reprint. Sorry about that. (Thanks to Gurupie Elliot Mosher for catching this so far in advance of the troubled week.)

Finally, I hope to have an NBA schedule posted by the end of today. There are some very significant differences by team in the early goings that will be critical to anticipate in your drafting. For example, in the first ten days of the season, Orlando plays 7 times, but Sacramento plays only twice. Suffice it to say that Chris Webber, Vlade Divac, or Jayson Williams should probably not figure in your draft plans! You can see the NBA schedule "sliced and diced" for any time period in the Sched-O-Matic. If you use a browser that is incapable of handling the Sched-O-Matic, I've posted just the output for the first 10 days so you can see what you're missing. Maybe you should upgrade to either Netscape or MSIE!

SW Football repricing continued to evidence the trading pattern we've come to rely upon. Players with a bye last weekend fared the worst. Evaluating the top 200 players in SWP/G, players with week-4 byes declined 7.4% in price. Players with a bye in week-3 increased by 2.2%.

The most unusual aspect was that players with an upcoming bye increased by 3.5%. Of the top 200 players, 22 have a bye in week-5, and 15 of those 22 posted price increases this week. Eight of them had increases exceeding $200,000. In the past, it was more typical for upcoming bye players to show much more mixed results, with high producers getting increases, but with the average reflecting a slight decline. Admittedly, with Washington in the mix, there were a good number of "high producers". Suffice it to say that the 3 teams with 10/10 byes will endure a pricing bloodbath next week.

There is also some evidence of "matchup trading" in this week's results. New England players appreciated an average of 7.1%, and that's coming from more than just their defense's attractive matchup vs. the Browns. However, with Cleveland facing Cincinnati this weekend, matchup trading may begin to shift to other weak sisters. It's hard to envision the Cincinnati defense as the favorite for next week. There would seem to be more attractive mismatches available. On the surface, Browns vs. Bengals doesn't even look like a mismatch.

CNN/SI prices once again went their separate way yesterday. You may recall that in the first repricing, SW and CNN price changes were slightly different. Then last week, the prices "mysteriously" converged. Yesterday, they moved apart again. (I'm sure glad I didn't take the early "head fake" and dismantle my separate CNN stats reporting!) Although the price changes again look pretty similar between the two game sites, there were four $1 million gainers at the CNN site vs. only two at the SW site. Four players showed CNN gains that exceeded their SW gains by more than $200,000: Hakim (+$280), Rhett (+$240), Emmitt Smith (+$240), and Biakabutuka (+$220). Four players showed CNN price changes that were at least $200,000 below their SW price change: Enis (-$320), Dyson (-$220), Manning (-$210), and Huntley (-$200). I can't see any particular pattern here, and it certainly could just be random noise. Anyone else have a thought on this?

I updated the SW football team rankings last night. Almost 10% of the RotoGuru universe now has a franchise value in excess of $70 million. The median price gain yesterday was about $5 million. With so many of the high priced running backs now sidelined, and with Kurt Warner producing so well at his low price, many teams are approaching the point that they can pretty much afford the team that they want. It looks like we are only a few weeks away from a significantly reduced motivation for gains trading. And we're just heading into week 5!

I also posted the final SW and CNN baseball team rankings last night. At least I assume they are final. (Please clue me in if that is incorrect.) I also discovered that the Echelon rankings that I posted yesterday were missing Monday's points, so I reran that as well.

If you haven't registered for Football Pickoff yet, come on down!

In fact, a new era begins today in RotoGuru-dom, as I introduce my first fully online-administered game, Football Pickoff. I'm using a new server for this venture, and I'm a little concerned about server responsiveness, but hopefully it will be able to cope. If it is sluggish, try again later. I admit the game site design isn't very artistic, but I think it should be simple to understand and navigate. And I probably built a trickier security (password) system than was necessary, but hey, I'm learning on the fly here, so bear with me. Frankly, considering this was all created within a little over a week, I'm quite proud of the first cut. And with all those caveats, I think you'll find the game intuitive and enjoyable, even if you only spend a few minutes on it per week - which is all it should require.

Hopefully, I won't have to endure too many technical snafus. I won't promise to offer a game that's trouble-free, but I will promise to keep you apprised of the situation if/when difficulties arise. So enjoy it! I know I'll enjoy producing it.

(By the way, if you registered during the game's "test drive" period, you'll have to start from scratch again. All of the pregame test data has been erased.)

I may be a bit tardier than usual in posting SW football price changes this afternoon. I have an early afternoon meeting that might run into repricing time. Give me an extra hour or so.

I posted final standings in Echelon's Ball Park Dreams baseball game last night. I'm still waiting for confirmation that SW and CNN points are complete before running the final rankings for those games. I'll coordinate the new Hall of Fame listings once all of the standings are in.

Now, I guess I'd better get started working on basketball draft stats. A RotoGuru's work is never done...

I'll wait until points appear to be final before posting the final team rankings. Echelon's are probably final already. Smallworld still has to catch up to 8/14, I think.

I always figured that in business, you should identify what you do best, and then exploit it as much as possible. Smallworld showed a great propensity to tick off its customers in the last two months, and just to drive home the point one last time, it would appear that they managed to have their baseball game on a server that was incapable of handling the last hours of baseball trading for each of the final three days. I suspect that the baseball game was moved to a lower capacity server some time ago (concurrent with their posted acknowledgment about slower stats) in order to get the other more active games on the higher capacity machines. But the final weekend was a real debacle for many. Three days in a row.

Segue to football, where Smallworld also has egg on its face today. The first round of stats posted yesterday included a double dose of points for players on some teams. I suspect their stats feed probably had duplicate entries for those teams, and if so, I'd just chalk up the problem to "one of those things happen from time to time." No big deal so far. In fact, soon after discovering the problem, they posted a notice that some points were incorrect and they were working to fix them. A good sign! We're not used to same day communication about problems! But then, several hours later, a notice was posted saying that points were now correct. As of this morning (10am EST), that message is still there. But the points are clearly NOT correct. In fact, players from at least four teams now show up with zeros. Since it is the same players that were originally double counted, it is hard to imagine how they could have escaped detection. Sounds like nobody at SW bothered to check. I'd say "unbelievable", but obviously, it's not. At least in football, they have a whole week to get the points right before the next set of games. But it certainly is a worrisome trend for basketball, where not only will points have to be updated daily, but now prices will have a daily opportunity for screwup as well. Could be a long hoops season.

By the way, Smallworld's latest server and quality snafus are especially poignant, coming directly on the heels of an email from Smallworwld's CEO extolling their technology upgrades and their new emphasis on Quality Assurance. Like I've said many times, talk is cheap. Promising to improve is easy. But the execution still doesn't demonstrate any commitment to follow through. (This is sounding like a broken record.) (Which also begs the question, how many of you have no idea what "a broken record" means in that sentence? Don't answer that! I feel old enough already...)

Congrats to Gurupie Matt S. who won the message board's football pool this week. It came down to a close tiebreaker with runner-up Rasputin. Both had the correct winner of the MNF game, each missed the margin of victory by 2 points, but Matt S. was 2 points closer in total points, so he got the nod.

All signs are go for a Wednesday launch of my new Football Pickoff game. Thanks to everyone who has helped me field test the programs. I have only a few minor changes to make today. So stop by tomorrow for the details. I'll also shift the baseball links to the bottom of the left menu tomorrow.

Echelon launched its basketball game yesterday. The format looks identical to their recent hockey game. You have $25 million to spend over the season. Each player has a daily salary which you are charged for each day the player is on your roster. There is no trade limit, but you do pay a charge to drop a player equal to one day's salary. Prices will change daily based on performance and demand, but roster salaries are locked in while you hold players. Points are allocated in the usual way, with the formula identical to Echelon's last basketball game. Should be interesting.

I'll have to wait to post final baseball standings. The biggest challenge this final weekend was not figuring out who would pitch the best, it was just figuring out who would pitch. And for SW baseball, server problems really caused trading headaches on both Saturday and Sunday mornings, from what I hear. I guess everyone needed to wait to see who would be pitching, and then jumped on the server all at once. I took my chances and traded earlier, and had no problems. No problems trading, at least.

My football stats source isn't yet available, so I'll be late with those. I'm hoping I'll have them sometime today, but I don't know what the problem is. I see Smallworld has already posted yesterday's points.

Meanwhile, hockey started this weekend. That doesn't affect the workload at RWH, but hockey enthusiasts should remember to take advantage of Paul Sheirer's excellent hockey site.

And Smallworld Hoops was launched this weekend. The scoring and roster formats look identical to last year. The big change will be a return to daily pricing. The is no mention of the daily price change cap, but prices will change at noon EST Monday-Friday. Trades will be doled out at a rate of four per week each Thursday. Aside from daily price cap considerations, this change seems likely to introduce more price volatility into the game, since there will be significantly fewer trades in each daily cycle, and thus the value of each individual buy or sell will be magnified. If managers still tend to blow most of their four trades each Thursday, then maybe the Friday change will show patterns similar to last season, with M-Th exhibiting lighter trading and more volatility. We'll just have to see. In any event, I'll be covering hoops again this year. With the introduction of my new football game this week, and the final processing of baseball stats, I probably won't have draft-related basketball stats up until next week. The NBA season doesn't start until November 2nd, so there's plenty of time. If you want to get started, though, the popular NBA Sched-O-Matic is already loaded with the full 1999-2000 NBA schedule, so you can begin thinking about the scheduling implications of your drafting.

Finally, I'm on schedule to launch my new Football game, which I've tentatively decided to name "RotoGuru's Football Pickoff". Game launch should be no later than Wednesday.

I watched a good bit of the Mets/Braves game last night. It was an excellent game. The Braves went ahead 3-2 in the 8th, and had the bases loaded with no outs, but failed to add any insurance runs. Then, with 2 outs and the bases empty, Edgardo Alfonzo hit a line shot over the left field fence to tie it. Both managers brought in their closers (Benitez and Rocker) to work the 9th, and they were both throwing gas. The Braves finally scored in the 11th after Shawon Dunston misplayed a fly ball into a triple.

So I believe the situation is that the Reds and Astros each need to win only once this weekend to at least force a one-game playoff - and that would still require the Mets to sweep the Pirates. Looking pretty bleak for Bobby Valentine.

It now appears that Kevin Brown will not pitch on Sunday. The Dodgers have announced Robinson Checo as the Sunday starter, since Brown cannot win 20. Meanwhile, if you're panicked that Saberhagen was AWOL yesterday, fear not. He was just pushed back to Friday, since the Red Sox have clinched. Even so, I wouldn't expect Sabes to throw many pitches. He's been on a short pitch count all season, and Boston plans to use him as their second playoff starter, so it wouldn't surprise me to see him go 5 innings and out today. Today's most consequential games, and probably the most popular starters, are the Reds (starting Neagle @ Milwaukee) and the Astros (starting Reynolds vs. L.A.). If you need to pick against the field, you should probably look elsewhere.

I'm still hoping to have my new "football winner picking" game ready for next week. Probably not by Monday, but I'm hoping for at least a "bare bones" launch by mid-week. The basic structure is that each week, each entrant picks the winners of any or all NFL games for the week. Points spreads will not be used. However, the scoring will reward the correct picking of an underdog more than the correct picking of a favorite. Here's how:

1. If a selected team wins, the points awarded will be 100 minus the percentage of entrants which picked that winning team. Thus, if a winning team was selected by 90% of all entrants, then only 10 points are awarded. But if the winning team is selected by only 10% of the entrants, then 90 points are awarded.
2. If a selected team loses, then it will lose points equal to the percentage of entrants which picked that team. For example, if the losing team was selected (to win) by 90% of entrants, then -90 points are "awarded". If only 10% picked that team, then only -10 points are awarded.
3. If an entrant picks neither team to win for a given game, then no points are awarded for that game.
4. If all games were 50/50, and all selections were random, then the expected total score would be zero. You get the greatest reward if you correctly pick underdogs to win. But, if you pick too many wrong underdogs, that can be costly, so indiscriminate underdog selections are not likely to be a good approach.

Meanwhile, I need a name for this new game. Any clever ideas?

Finally, thanks to Gurupie J. "Turd'L" Miller for suggesting today's quote.