Go forward to more recent blurbs.

1/29 - A minor bug - a gnat, so to speak - was detected in the Assimilator yesterday. The last player on a roster was incorrectly being recalled when read from the cookie file. It's fixed now. I'll let you decide whether I put that one in there intentionally just to see whether anyone was paying attention. If I did, some of you passed.

SW Hoops teams for the RotoGuru rankings continued to arrive yesterday. Let me reiterate a point I made last week. You need to send me your team name and user_id  even if you are in a "RotoGuru" division. The division look-up routine isn't showing team names, at least for now, so I can't find them on my own. This may also preclude me from doing any division rankings this season. (Not a big deal, if you ask me.) Check the listing of teams I have received so far, and if you're not there, shoot me the requested info.

A few more notable performances eyeballed from yesterday's NBA preseason boxscores:

• Milwaukee: Allen 31.5, Robinson 22.5
• Minnesota: Garnett 51, Marbury 23, Mitchell 21
• Portland: Sabonis 34, Williams 32.5, Wells 22.5, O'Neal 20.5, Cato 18
• Seattle: Baker 28, Williams 24, Payton 17.5(no assists!), Polynice 16
• Sacramento: Webber 45.5, Divac 42, Abdul-Wahad 21, Williamson 21
• Golden State: Mills 30, Cummings 25, Dampier 22.5, Marshall 20 (0 pts), Caffey 17

1/28 - Amazingly enough, no one's found a bug in the new Assimilator yet. Maybe I should have planted one in there, just to see if anyone has even tried it out. (Actually, a few have emailed me to say it worked fine for them.)

Have you been looking over the first few preseason box scores? I have, and so far, very few guys look like they are ready to produce. I think the only player who put up better than 40 SWP in a game was Charles Barkley, with 40.5 SWP (in spite of 7 missed free throws). Other notable players' SWP results:

• Houston: Pippen 36.5, Olajuwon 28
• San Antonio: D. Robinson 28, Duncan 19.5
• Indiana: D. Davis 30.5, Jackson 18.5, Miller 17, Smits 16
• Chicago: Harper 28.5, Rogers 16, Kukoc 12
• Toronto: Christie 28.5, Willis 17.5
• Boston: Mercer 37.5, Anderson 33.5, Antoine Walker 30.5, Paul Pierce 31.5
• Cleveland: Ilgauskas 33.5, Kemp 31.5, Knight 25.5, Person 15
• Detroit: J. Williams 34, Hill 27.5, Vaught 26.5, Stackhouse 25, Reid 23.5
• New York: Ward 37, Childs 32.5, Sprewell 32, Houston 19, Ewing 18.5
• New Jersey: Cassell 26, Murdock 25, McIlvaine 22, Gill 21, Jayson Williams 18.5, Gatling 17

1/27 - Today I introduce a "beta" version of the new and improved Basketball Assimilator. Here's a list of the primary new features:

• Accommodates both Smallworld Hoops and Echelon Full Court rosters and scoring (there's a "toggle" button that allows you to indicate which type you're entering)
• On the stats page, a column has been added to indicate the player's opponent for the most recent day. This helps you identify whether a zero resulted from an day off, or from an "off day".
• Below the stats table is a new table that shows a day-by-day schedule look-ahead for each player for the next 10 days. This should help you anticipate possible trade opportunities related to scheduling.

1/26 - As I was doing some tinkering with potential drafts for both the SW Hoops and Full Court games, I was struck by the dramatic difference in the relative affordability of talent. For a $50 million SW team, you can get 10 players that average roughly 20 SWP/G. For a $15 million Full Court team, you can draft 12 players that would average more than 30 SWP/G. (Although Full Court uses a different scoring formula, I'm representing the output in SWP to allow for comparability.) The team I'm considering drafting in Full Court would cost more than $100 million to draft as a Smallworld team. (Actually, it would cost $125 million, but I've factored it down since a Full Court roster can have 12 players, while a SW roster has only 10.)

I raise this comparison not to suggest that one game's approach is preferable to the other. I just thought the distinction was striking.

Perhaps an even more important difference in the draft pricing of the two games relates to rookies. Smallworld has obviously done some significant handicapping of rookie performance. While the average price of a Smallworld rookie is only about $3.5 million, the average price of the top 25 rookies is $6.7 million, and seven rookies have 8-digit SW prices. In Full court, there is very little differentiation among the rookie prices. In Smallworld, the average rookie costs about 65% of the average returning player, but the average top-25 rookie costs about 130% of the average returning player. In Full Court, all rookies are priced at approximately the same price as the average returning player. This suggests that an attractive Full Court strategy would include drafting a number of the most likely high-performance rookies. And if you're wondering who those rookies might be, you can look at the handicapping implicit in SW's pricing for their outlook.

For example, the most expensive rookie in the SW game is Mike Bibby. His $12.3 million price tag is comparable to the likes of Mookie Blaylock, Doug Christie, Brevin Knight, Ray Allen, and Mark Jackson - all who average in the vicinity of 30 SWP/G. In Full Court, Mike Bibby is priced at $817,500. Those same five players in Full Court have an average price of just under $1.6 million, almost double the price of Bibby. This is not to suggest that Bibby will perform at the level of those five veterans. But if he does, he sure represents striking value. So much so that one possible Full Court strategy could be to draft an all rookie team, and then after a few weeks replace the laggards with proven veterans.

It's something to think about. Certainly more interesting than the Super Bowl drivel.

1/25 - I've been thinking about the trade value of cheaper players in the SW Hoops game. Since managers will have to buy some low priced players - particularly if they also want to own some superstars - will low priced players get more early season trading activity? And if so, are they likely to produce decent price gains, if these players produce at a level better than the cheap price indicates?

Experience has shown that the best trade values in early trading preiods are those players on Smallworld's top 50 player lists whose prices look abnormally low relative to their peers. These lists are the only source of information on relative SWP production that many managers see. So, what will it take to land on those leaderboards?

I did a quick simulation of the way the leaderboard should look on Tuesday, Feb 9th - the day prior to the first repricing. (This assumes that the top 50 player lists will be operational by that date - not at all a foregone conclusion!) If all players produce at their prior year averages, the cutoff point for making the top 50 list for all positions combined will be around 75 SWP. For guards and forwards individually, the cutoff will be closer to 50 SWP. And for centers, it will be much lower - perhaps 10 SWP or so. Using this criteria, and considering the actual schedule through 2/8, only one player worth less than $2 million would appear on the overall top 50 list. (I'll let you figure out who it is - but he's not tough to find, if you use my stats tables.) On the position-specific lists, no guards under $2 million would appear, 2 forwards under $2m would be listed, and 11 centers would make it. Remember, this all assumes that players produce each game at their prior year average - thus, rookies are ignored in this analysis.

The point I'm leading up to is that, unless a cheap player is grossly underpriced (and can produce at a pace of 20 SWP/G or better), he's going to be tough to notice. Centers may be the exception to that rule, though, as their performance threshold for listing will be much lower. Most players who averaged around 20 SWP/G last year are draft-priced around $8 million. So, when drafting your cheapies, it's probably best not to expect too much from the guys who are only modestly undervalued. They are unlikely to attract big buying.

Also, since the high priced players are unlikely to generate big gains early in the season (given their lack of affordability), a barbelled roster of "stars and scrubs" may have a tough time showing much in the way of price gains. You're more likely to find those gains coming from players who are priced nearer the $5 million average - especially those who are producing well enough to land on the top 50 lists. Some of these players may also benefit from heavy scheduling early in the season - remember that SW lists players by total points, net per-game averages. So players with extra games played in the early weeks will have an artificial advantage in the rankings. When it comes to trading patterns, perception is reality! Forewarned is forearmed. 'Nuff said.

Yesterday afternoon, I (finally) started working on modifications to the Assimilator for basketball. I probably won't have a "beta" version ready for several more days. But I will be incorporating several new functions. First, it will accomodate rosters and scoring for both Smallworld Hoops and Full Court games (I'll put a button in it to allow you toggle from one to the other). Also, I plan to include some schedule look-ahead data for each player, so you can easily spot players with heavy vs. weak schedules. Incorporating these items will take a few days, but with the season still 10+ days away, I'm not feeling any pressing sense of urgency.

Finally, I'm feeling a bit uneasy about the completeness of the player-team affiliations in my stats tables. Although I'm checking three different sources to identify trades and free agent signings - NBA.com, ESPN, and RotoNews - I'm finding that each of these sources lists transactions that are missing on the others. Then I read in the Hartford Courant this morning that Alan Henderson signed with the Hawks over the weekend, but so far, this hadn't yet been picked up by any of the three internet sources. So I checked the web site for one of the Atlanta papers, and sure enough, it reported that Henderson had signed a new seven-year contract on Saturday. This is a long-winded way of saying that, if you find players that I'm listing as free agents, and you think they have already been signed by some team, please let me know. I think I'm going to need all of your eyes and ears to get these team affiliations correct before opening day. Thanks.

1/23 - Look around! I've got something to show for yesterday's work. For example, I updated all of the player/team affiliations in the stats tables for the recent flurry of trades and signings - but you'll still notice a lot of unsigned free agents. I actually started from scratch - filled in the status of every player, based on info at NBA.com. Since the process was entire manual, some errors are likely - so please report anything that looks wrong. Hopefully, we'll converge on the truth by the opening day.

But my piéce de resistance is the color-coded NBA schedule. I developed something like this for myself last year, and while it was somewhat helpful as a means to spot teams that were heavy or light with games, it wasn't as actionable, because trades were generally too precious to use just to pick up a few games. But this year, we've got more trades, less time to spend them, and a schedule that looks pretty erratic from team-to-team. This new schedule format aligns all 29 team schedules side-by-side, and indicates by green or red shading whether the team is in a period of heavy scheduling (4 games in 5 days, or 5 games in 7 days), or light scheduling (1 game in four days). By using the color, it's easy to pick out teams that are worth looking at - in either direction. The downside to the table is that it's a pretty large file to read - almost 150kb. If you have a slow modem, you might want to work with the abbreviated 4 week schedule, which is slightly under 50kb. As each week passes, I'll lop off the front week and add one at the rear. I hope these will help you plan ahead so you can time your roster moves as efficiently as possible. As far as I know, it's the only schedule of its kind on the web - a RotoGuru original!

I may also try to put some sort of game look-ahead count on the player stats tables, but I haven't yet figured out what would work best. I'm open to suggestions.

1/22 - Yesterday turned out to be a day of administrative activities, including updating some of the banner links I have at other sites. Consequently, I didn't get much done in the way of further hoops analysis, nor did I get started on modifying the Assimilator for Hoops. I hope to have a preliminary version available for "test driving" early next week.

I did take a quick look at the just-released NBA schedule last night, though, and it looks like we'll need to pay close attention to it when picking our draftees and planning our trades. For example, there are five playing dates prior to the first repricing on Feb. 10th. During this initial 5-day period, seven teams play four games (Atlanta, Boston, Charlotte, Chicago, Detroit, Miami, and San Antonio), while six teams play only twice (Clippers, Milwaukee, NJ, NY, Portland, and Vancouver). The other 16 teams play three times. Then, the following week, 12 teams play four times, while two teams play only once (Boston and Washington)! I'm going to try to figure out a format for posting the entire schedule which will help you manage this aspect. More on this next week.

I want to continue the recent practice of RotoGuru team rankings for the Smallworld Hoops game. Unfortunately, for some reason I can't get at the list of teams in the various RotoGuru divisions. I know there are at least 19 divisions with names that start with "RotoGuru", but I can't see any of the team names in those divisions. So, if you want to be included in the RotoGuru rankings, you'll have to send me your team name, user_id number, and game site (SmallWorld or CNN/SI) regardless of whether you're in a RotoGuru division or not. I'll start taking your submissions today, so send 'em in. This will help me spread out the preparation work, so I can hit the ground running once the season starts. (If you need help finding your user_id, see the instructions at the bottom of the Football team rankings page.)

Click to subscribe to "Gurupies"

From time to time I've gotten Gurupie requests to start a mailing list service, for which I would email all subscribers a copy of new strategy articles as they are posted, plus weekly summaries of blurbs. The purpose would be to cater to those of you who sometimes just don't get around to checking the site everyday, to make sure you were alerted to new strategy articles, and to remind you to stay tuned to my "blurb thoughts". So let's try it on an experimental basis and see how it goes. To facilitate administration, I have set up a mailing list at onelist.com, called "Gurupies". As I post new articles, I'll send a copy to everyone who has registered for the list. I'll also send out a weekly blurb update. Hopefully, most people can receive email in html format, because it will greatly simplify things if I can just send the html pages (or an extract) verbatim. If you're interested, click on the OneList button, and register. It's free, and this will allow you to control whether you want to remain on the list, switch email addresses, etc. We'll see how it goes. (If you visit the RotoGuru site daily, then don't bother, because all emailed material will be available at the site first.)

Now, let's see if I can figure out who's on what team. . . .

1/21 - I posted by first Hoops strategy article yesterday afternoon. Same Game, Different Feeling" examines some of the implications of the new pricing structure. While it will take some getting used to, I'm generally positive on this change. As always, I'm interested in your feedback.

Smallworld changed the listed positions for nine players yesterday, in each case putting the player at a more logical position. The nine players (and their new positions are:

Pos  Name
F    Abdul-Wahad, Tariq
F    Bennett, Mario
F    Fortson, Danny
G    Graham, Greg
G    Jackson, Bobby
G    Legler, Tim
G    Long, John
F    Mercer, Ron
F    Wingfield, Dontonio

1/20 - I've finally started receiving some more thoughtful analysis of the implications of the rules and relative pricing changes for the Smallworld Hoops game. I posted the first Hoops-related feedback letter late yesterday, and hope to have more to relay in the coming weeks. Early feedback was generally griping about the inability to afford a "reasonable" roster given the extraordinarily inflated prices. Frankly, though, I think the price adjustments have some merit, and will probably benefit knowledgable NBA enthusiasts at the expense of more casual observers. Secondary NBA players who have generally been irrelevant to the SW game will now assume a much more critical importance. And users of the expanded statistical data provided at this site will probably be at a greater relative advantage than those who rely solely on the SW site for their information.

RotoNews links for each player name were added to all stats tables yesterday, facilitating your ability to "drill down" from the stats tables to see what the latest word from RotoNews is. Unfortunately, the status of many NBA players is still in limbo, and with the lockout still in place, the schedule still unannounced, and rosters still highly uncertain, it still feels like what we don't know oftens overwhelms what we do know. But now I see Feb. 5th mentioned as the season opening date. (Early league announcements suggested a Feb. 2nd start.) So there is no need to have your roster set for another couple of weeks. (Smallworld has said that trading limitations will begin after noon on Feb 2nd, but I suspect that may be relaxed until the eventual opening day.)

If I can maintain my attention span, I'll have my inaugural "Hoop Pointers" article ready for tomorrow, discussing the rules and pricing changes and my take on their strategic implications. So, I'd better shut up now and get to work.

1/19 - The team rankings were updated yesterday afternoon, and lo and behold! I had the best week #3! (In fact, I cleaned all your clocks - beating the second best team by 116 SWP!) If only I hadn't screwed up by holding on to Miami in week #2. In addition, my oldest son's team is now ranked #20 worldwide, and second in the RotoGuru rankings. So, the high fives were flying last evening in the Hall household. Glad I didn't throw in the towel after the second week.

Steve Houpt has again submitted the best 3-week roster score so far, a total of 5717 SWP. (Too bad his real team isn't living up to his after-the-fact prowess!) Steve just built off of the same roster that produced the best 2-week score, dropping Marino, Curtis Martin, Adrian Murrell, Jimmy Smith, and Keyshawn Johnson, and picking up Chandler, Terrell Davis, Harold Griffith, Matt Hatchette, and Terance Mathis. I haven't tried to gear up my computer program to assess the possibilities, but I might try that later in the week. Too many basketball projects for the time being, though.

Before turning to basketball, though, I stumbled across a contest, sponsored by the Pro Football Hall of Fame, to pick the list of 1999 enshrinees. Click on the link above if you want to check it out. I have no affiliation with this contest - just thought I'd call it to your attention. No extra charge. (If you win the contest, send me a postcard from Canton next summer.)

I added a set of sortable stats tables for the Full Court game yesterday. The next projects will be to add links from player names to their RotoNews information (probably later today), and to begin to try to get players assigned to the right teams - which may take awhile. I also hope to have a basketball Assimilator ready later this week. And, I want to begin to do some analysis.

1/18 - First, football. Good games yesterday. I fared pretty well, with both Chris Chandler and the Denver defense being my key picks for the week. Now we come down to just two teams, and the roster choices are pretty limited. Not much strategy left. Just get your rosters set, and wait. Or better yet, wait for 12 days or so, and then get your rosters set. No sense getting blindsided by a freak injury in the interim.

I'll probably have the updated team and division rankings posted by tomorrow. I'll also want to update the analysis of the best possible score through three weeks - but that can wait for a few days.

Now, on to Hoops. You have probably already noticed a few cosmetic changes to the left menu. I added the various ball icons to the left of each menu item to make it more obvious which sport the links relate to. Hopefully, this will make it less likely that you'll end up in the wrong sport. I've introduced tables for Smallworld Hoops, including stats using this year's formula applied to last season's stats (regular season only). You'll easily detect the formula that SW used to determine draft prices. And that will also help you see those players whose prices were adjusted. Pay particular attention to the cheaper players - because you'll need several of them. Maybe more than several. I still want a few days to digest the implications of the inflated prices and the new trade rules, so I'm going to reserve judgment for now. Draft analysis will be tricky as well because so many players are still unsigned, and because the schedule still hasn't been announced. But we have plenty of time to kick around strategy ideas before those drafts have to be finalized. The season doesn't start until Groundhog Day - 2 days after the Super Bowl.

I also hope to add some statistical coverage of the Echelon Sports'Full Court game this season. I'll probably just clone the Smallworld tables, but use points and prices from the Full Court game. I hope to have the first cut of those tables done within a few days. Ditto for the Assimilator. Meanwhile, check out some of the other new pages, particularly the favorite links. Let me know if you find glitches (I expect there'll be a few), and if you have any additional links to suggest. I'm not looking for an exhaustive set of links. Just the best.

1/16 - A rare weekend blurb!

Two noteworthy events: First, I bit the bullet and wrote a simple, brute force program to find the best set of plausible rosters for the first two weeks of the playoff game. After running continuously for 48 hours, it looked like the Excel Solver routine wasn't going to get there. I suspect it was the trading constraints that gave it problems. So, with a little bit of ingenuity, I wrote a simple, brute-force model that solved the problem in just under 90 minutes. It considered almost 55 trillion roster possibilities, and fully evaluated 129 million of them (the rest were dismissed before a full evaluation was necessary). Bottom line - Steve Houpt had it right all along. 4228 SWP is the best possible performance through two weeks. I wonder what next week will bring?

Second item: Smallworld Hoops was introduced late yesterday. RotoGuru divisions are already forming. I haven't had much chance to evaluate the rules and pricing, but it is clear that some strategic adjustments will likely be warranted. In particular, draft prices are much higher relative to the team cap of $50 million, which means that you'll need a number of "scrubs" to complete a valid draft roster. No more "all star" rosters. More strategy analysis to follow next week. Preliminary price and stats tables will be available sometime next week as well. (No promises on the day yet.)

Enjoy the weekend!

1/15 - My playoff roster optimizer program is still cranking away. After 38 hours, it still hasn't found the optimum two weeks of rosters. It appears to be closing in on the right neighborhood, but I can't tell for sure. At this point, I think it would have been faster to have it just test all of the possible player combinations. After narrowing the list of eligible players to about 40, I figured there are roughly 15 million possible combinations of players to produce a properly configured roster by position for a single week. Not all of these are affordable, of course. But if you add a second week, the number of rosters to be tested jumps to about 218 trillion (give or take a few). I could have set up a routine to cycle through all 218 trillion combinations, testing first to see whether the price caps are satisfied, and then to see whether the trade limit is satisfied, and if so, what the SWP total is. But I figured it would be prohibitively long to run this. Now I'm less certain. In any event, at this juncture, I'm more interested in the mathematical process than the result. A classic Guru personality "quirk".

Anyway, this weekend will produce at least one NFL "first" - as it will be the first time a "dome team" will advance to the Super Bowl. Rob Neyer authored an interesting analysis of this at ESPN's site this week, although I think he missed commenting on an important consideration. His primary rationale as to why a dome team hasn't made it into a Super Bowl is that, in general, dome teams that even made the playoffs haven't been that strong (in terms of regular season won-loss records). What he fails to explore (or even suggest) is whether being a dome team has something to do with that lack of success in the regular season. It would be interesting to see historical regular season records of dome teams, broken down by home vs. away, and for away game, by those played in domes vs. those played outdoors. However, I'm not so interested in it that I'm willing to do it myself.

The first RotoGuru-recommended free NBA fantasy game has emerged, from Echelon Sports (formerly Central Plaza). Full Court Basketball is now accepting registrations. Three RotoGuru leagues have already been started, with league ID#s of 104, 111, and 152. In Central Plaza's Touchdown Fantasy Football game, all five RotoGuru leagues finished in the top 10 of all leagues, ranked by the average score of the top five teams in each league. The grand prize winner was also a Gurupie. First prize for the basketball game is $1500! I haven't fully digested the rules or strategic nuances of this basketball game, but it appears to be a pretty straightforward game with a hard salary cap (which escalates gradually over the season, unlimited trading (subject to a 4-day minimum holding period for all players), and a formula point scoring system. In addition, Gurupies who have played previous Central Plaza games have been favorably impressed by the smooth operation and minimal number of technical glitches. After the first three leagues fill, I'll be happy to post league ID#s for newly created RotoGuru leagues as well.

Next week I hope to begin full coverage of fantasy basketball, with stats, links, recommended games, and strategy analysis. Be sure to check in.

1/14 - Gurupie Frank Kovacs sent me the following email yesterday:
"How could I have gone up over 6100 spots on only 1,249 points, and be ranked #658 worldwide? The best possible total for each week is well over 2,000 points. Now, there was one guy in my division that did not make any trades for the second week. He ended up losing 302 points! Do you think this could be the reason? Maybe people who originally set their teams for week 1, missed week 2? Did you see a lot of this in your ranking?"

Frankly, I hadn't paid any attention to ranking changes, or point totals for the second week. (I was after all, pretty disgusted with my own Miami-induced 473 SWP showing, weak enough to drop me about 6000 spots in ranking.) But Frank's inquiry sent me off to look at this some more, and I found some surprising results. In fact, I added a column on the team rankings that shows the change in world rank for the week, so you can see the relative results. Here are some of my observations:

• First, take a look at the total point increases for just the second week. In the team rankings, the best point total was for a team which started fresh after the first week, so he was unconstrained by trading considerations. Other than that team, no team cracked the 1,400 SWP barrier, and only six teams broke 1,300 SWP. So Frank's total for the week of 1,249 looks pretty good on a relative scale.
• Most teams with a 4-digit week #2 showed a bigger jump in ranking than Frank experienced. Frank's limiting factor was that he started relatively high to begin with.
• The principal culprit in the shoddy second week totals seems to be team defense. Roughly one-half of the RotoGuru teams kept the same defense from week 1. But many of the switchers did so because their week #1 defense was eliminated. If you look at the teams which had Jacksonville or Arizona for the first week - the two best defenses by far - almost 70% of those teams kept those same defenses for week #2. And paid a heavy price. (Looks like I had a lot of company.)
• The top 50 worldwide teams after week #1 had similar problems. Only five of the original top 50 stayed in the top 50 for week #2. And those five teams only averaged 1,135 SWP in the second week! Of the other 45 teams, one must have had an invalid roster for the week, while the other 44 averaged only 549 SWP in week #2. Whoa! And again, the primary culprit was retaining the week #1 defense, notably Jacksonville. (One of the top 50 teams remained in the top 50 while keeping Jacksonville, however. Nice offensive picks!)

1/13 - A few submissions for the best possible rosters for the first two weeks cam in yesterday. The clubhouse leader after one round is Gurupie Steve Houpt, who figured out a set of plausible rosters which would have produced 4,228 SWP, more than 900 SWP better than the #1 worldwide team. To see Steve's submission, click here. (Note: I've put his rosters on a separate page, along with the points and prices for all players, since I realized many of you probably wouldn't have the data organized in one place. I also set the link to open a new browser window, so you can open it without losing your place here.)

Here are Steve's observations on his submission:

• "One of the biggest things I see in my best first two week entry is that just because a player or team makes it to the second round, it is not necessarily in your best interest to keep him. (11 playing slots is not always the best) History shows that 75% or more of the first round teams will lose in round two and you will have to trade the player the next week anyway.
  1. Selling Fred Taylor allows you to buy Cunningham and Martin.
  2. But most important, and I think you expressed this yesterday, chances are it's not worth keeping a first round defense. In your long range plan, count the first week defense as gone, a must trade. (i.e. - keeping a first round defense to fill Jeffers would have been disastrous). Looking at the first week's top 50, 44 of them kept first round defenses and most are gone.
  3. Easy in hindsight."

1/12 - With just two weeks and four teams left, the player choices are fewer, but the management strategy alternatives seem almost as complex. Do you position your team to benefit from the favored Super Bowl matchup of Denver/Minnesota, or do you look to differentiate your team somewhat by picking an underdog? Do you pick the consensus affordable players, or pick some cheap players to be able to afford someone like Terrell Davis? Are you trying to catch up to competitor who is ahead of you, or maintain a lead? Are you still scrambling for a top 50 worldwide ranking, or are you just trying to avoid an embarrassment?

This will probably be the slowest week at RotoGuru for some time. Although next week starts the two-week hiatus before the Super Bowl, it will probably also bring the first opportunities to prepare for the shortened NBA season. In fact, I'm not sure I've got much to add for the rest of this week. But for those who like a good challenge, read Steve Houpt's latest feedback, and then my response. The question is, "What is the best possible result for the first two weeks?" It's not too difficult to figure out the best possible roster week by week, but when you need to link those weeks together with no more than five substitutions, it's a much more challenging exercise.

So here's the deal. For you Gurupies who like intellectual stimulation, let's see if we can collectively develop the ultimate roster(s). So far, we have only two weeks to deal with, so this week - let's figure out the best possible point production for the first two weeks. I haven't tried it yet, but one place to start is to look at the best actual team, which has already amassed 3302 SWP. If you can figure out a set of $50 million rosters for the first two weeks (which differ by no more than five players), send me an email with the two rosters and the total points for the first two weeks. The Assimilator should be useful in tallying the points for each week, although you'll have to store each week's roster separately. You can also simplify the email process by saving the calculation page from the Assimilator (after you click on the "calculate roster points now" button, save the output as an html file using the browser's "save as" feature, and then attach the two html files to your email. Each day, I'll post the best submission so far, as well as the name of the submitter. But if you can't beat 3302 SWP, don't bother. Not only might this be in interesting computational exercise, but it might even provide some insights on the best strategies. This is the second SW playoff game I've played (they had a Hoops playoff game last spring) with a similar framework, and I still think I don't appreciate many of the strategic nuances - though I'm learning. (One of my favorite concepts is that you never really understand how to do something right until you've first screwed it up a few times. Miami's performance this weekend made a lasting impression, no doubt.) But while screwing it up in practice, it might also be instructive to model how the game could have been "done right", using the benefit of 20/20 hindsight.

Plus, as I said, it's a slow week. Your 15 minutes of fame awaits.

1/11 - Time for damage control!

I knew I was taking a risk going with the Miami defense this weekend, but I decided to brave them for a few reasons:

• They were my first week defense, so I could save a trade by keeping them.
• They had handled Denver well in November - albeit in a rather meaningless game for the Broncos.
• There was a good chance of cold weather in Denver, which would presumably hurt the offenses.
• I had Elway, so as long as he was a key factor in Denver's offense, I was hedged.
• Miami didn't look like a common pick at defense, so if they had a surprisingly good outing, it would be a competitive coup.

1/8 - Not much left to be said. If you're still looking for more stats, see Steve Houpt's latest feedback letter. Then get your week 2 rosters finalized, and enjoy the weekend.

1/7 - Frankly, if not for the fantasy sports opportunity it affords, I would have been quite satisfied to see the NBA season torched. That's not really a fair comment, however, because I would have said the same thing last season, too. Even now, I must confess to being somewhat ambivalent about the settlement. Sure, it will provide something to keep RotoGuru.com "in business" during February, and that's good - I guess. But March would have been busy anyway, with March Madness and baseball drafts, and then baseball starts in April, so the NBA season will be "the only game in town" for only about a month. I was starting to look forward to a February vacation.

SW will be announcing their Hoops format sometime around January 18th. I expect Central Plaza will also host an NBA game. I have no idea about MLM - they have been busy retooling their site, and have effectively shut down the football game since late November. (A beta version is currently accessible at their site, but it's not fully functional.)

So I'll now have to dredge up my stats from last season, and prepare to defend my 1998 SW Hoops title. I'll try to have some stats ready around January 18th as well, though it may take a day or two to accommodate any new scoring changes in the SW game. If you're looking for a quick and "easy to import" source of 1997-98 NBA stats, though, I recommend those at sports.yahoo.com/nba/stats/. You can use your mouse to copy and paste the stats into a spreadsheet, or you can even download a more detailed, comma delimeted text file at bball.yahoo.com/nbastats/stats.txt.

Meanwhile, as promised, I took a peek at the top 50 worldwide teams in the football playoff game. Not surprisingly, every single team had either Jacksonville or Arizona for its defense last weekend. Many of these teams haven't done their second week trading yet, though, so a current snapshot shows a mixed bag of old and new rosters. I did note that two of the teams have 10 week one slots remaining active for week two. For all 50 teams, the average number of active players carried over from week one is 7. Of course, unless at least one underdog team wins this week, all of those rosters will need massive retooling for week 3, so I'm not convinced that they have as significant an advantage as it may appear. The most interesting find was the kicker for the team ranked 22 - Chris Boniol. Obviously, he's not going to score any points (though that still places him ahead of Richie Cunningham), but his price tag was only $720,000, making him the only kicker under $1 million, so it frees the most cash for the more prolific positions. Interesting.

1/6 - Team and division standings for the playoff game were posted late last night. Only one RotoGuru team cracked the top 100, and that team only barely made it, at #94. Better luck next week, Gurupies. In tomorrow's blurb, I'll take a summary look at the rosters of the top 50 worldwide teams, so you can compare against the early leaders.

BTW, for the team rankings, I picked up all teams from the RotoGuru Divisions, plus all other teams which were submitted to me via email. Teams with total first week points of zero or less were ignored. If you want your team or division to be included in the rankings, follow the submission instructions at the bottom of the related rankings page. And if your team should have been included but I somehow missed it, sorry about that. The collection process is semi-manual, so some errors are likely. Please submit again.

Here is the breakdown of the 237 RotoGuru team defenses for the first week: :

Defense          % held     SWP
Jacksonville      45.1%    -151
Buffalo           19.0%    -537
Miami             15.6%    -331
Dallas             7.6%    -456
Arizona            4.6%    -155
Bye team           3.0%    -700
New England        2.1%    -538
San Francisco      2.1%    -483
Green Bay          0.8%    -492

1/5 - In response to multiple requests, I've calculated the best roster for the first weekend subject to a $50 million salary cap. The best I could produce was 2,292 SWP, which I posted directly beneath the best roster with an unlimited budget. If anyone can produce a better result, let me know and I'll give you credit.

I reset the poll question this morning. While the total votes cast in the last poll was impressive - over 60,000 - I suspect that one or two users figured out a way to cast votes in an automated repetition. So I don't think there is anything valid to read into the results (if there ever is about any of these polls). So today's poll no longer allows for repeat voting (at least, not from the same computer using the same browser. The poll software uses cookies to determine of you've voted in this poll before.)

Regular season team and division standing were updated yesterday afternoon. As pre-announced last week, the best RotoGuru team was "BLB #2", managed by Brian Butvin, of Cleveland. A Tribe fan and a computer software developer, Brian has been playing fantasy sports for five years - starting before the internet made administration and availability so easy. Brian says this season - in which his team placed #17 worldwide - has been his most impressive result so far. How'd he do it? Here's Brian's own assessment:

• "My team was put together after I decided that quarterbacks and running backs were the most important positions on the team. I initially chose the two quarterbacks who I believed had the best receivers. That would be Steve Young of San Francisco and Brad Johnson from Minnesota. For running backs, I went with all-round backs like Marshall Faulk, Curtis Martin and Ricky Watters. I went cheap for starting wide outs, picking Charles Johnson, Terry Glenn and Curtis Conway. My kicker was John Carney. My tight end was Irv Smith (the best available for the money left). My biggest conceptual mistake was to go with the cheapest defense. I felt that defenses didn't amount to much and the added money allowed me to take some nice mainstream players. I got lucky because Oakland turned out to be one of the best defenses through half the season! (By the way, I picked this team in only 29 minutes atthe last minute - what the heck, I thought!)"

1/4 - Ignorance would have been bliss!

If I hadn't been paying attention last week, I would have had Garrison Hearst and Eric Moulds on my roster for the weekend. However, when Hearst's injury looked to be problematic, I decided to play it safe, and dropped him in favor of Emmitt Smith. Then, to square up my roster, I swapped Moulds for Terrell Owens, plus a few other minor trades. As it turned out, Hearst and Moulds did MUCH better than Smith and Owens. In fact, had I just left my roster alone, I'd have had 1,748 SWP for the weekend, which I'm pretty sure would have been a very good result. Instead, I got only 1,322. So the trades cost me more than 400 SWP. Argh!

In any event, the Assimilator is now updated, and the points for the weekend games are also posted on the weekly SWP page. I had some travel difficulties getting out of Cleveland over the weekend, so I'm a bit behind in some of my other processes, which need some minor tweaking to make them "playoff friendly". (My travel problems caused me to miss the end of the Buffalo/Miami game, and the entire Green Bay/San Francisco game as well. Bummer!) I also have some other non-RotoGuru errands to run today, so I'll be getting the rest of the site current over the next couple of days. Here's my tentative plan:

• Sometime this afternoon, I should have updated summary stats tables posted. By this evening, individual player stats pages should also be updated.
• Sometime today, I'll download the final results from the regular season game, both the team rankings and the division rankings. These will either be posted this evening, or tomorrow morning at the latest.
• Once I get the regular season rankings posted, I'll make the necessary modifications to capture the playoff team and division standings. These probably won't be posted until sometime on Tuesday. I will pick up all teams in RotoGuru leagues, plus all other teams and leagues that Gurupies have emailed me.
• After the team rankings are completed, I'll announce the winning manager, and post his name in the Hall of Fame. This may not happen until mid-week.
• For now, I'll just leave the current poll. It's obvious that a lot of people have been "ballot stuffing", which was allowed in this poll, partly because I was interested in the full ranking (rather than just the top vote getter), and partly just to see how things would progress. At one point, I knocked 1000 votes from the Denver/Elway vote total, because it was obvious that someone was going wild with that selection. But later on, I restored the votes because the other categories caught up. Obviously, there has been some Yankee stuffing as well. I doubt that future polls will permit repeat voting, but since I don't have the next poll question ready yet, and I've got plenty to do, I'll just leave this one up a day or two more.

1/1 - A few last minute comments before you finalize your playoff draft:

First, be aware that six additional players were added yesterday, including Jonathan Quinn, Chris Jacke, Scott Zolak, Terry Kirby, Dave Meggett, and Matt Hatchette. I've added these players to the summary stats tables and to the Assimilator. I do not have accurate game-by-game histories for them, however. My game history database wasn't transferred to my laptop. I get them updated next week.

Second, make sure you're up to date on the injury outlook. Last I heard, Zolak was definitely starting, Brunell was probable, Abdul-Jabbar was probable, Hearst was still questionable, and Deion was doubtful. You have until Saturday morning to get your roster set.

Third, as I thought about the game a little more, it seems like the first two weeks will probably be the most important ones. As the field of teams shrinks, rosters will necessarily converge. So the best chance to get ahead of the pack will be at the outset.

Enjoy the weekend.

Click here for prior daily blurbs, by month:

December . . . . . November . . . . . October . . . . . September . . . . August . . . . . July . . . . . June . . . . . May . . . . . April . . . . . March

RotoGuru is produced by Dave Hall (a.k.a. the Guru), an avid fantasy sports player. He is neither employed by nor compensated by any of the fantasy sports games discussed within this site, and all opinions expressed are solely his own. Questions or comments are welcome, and should be emailed to Guru<davehall@home.com>.