November 20th, 2006
|10:11 pm - Major League Chess revisited, plus general topics of sports organisation|
Just over a year ago, I blogged about the start of the US Chess League. A little over a year later, the league is flourishing and about to conclude its second season. Time to look again at the US Chess League enterprise and see whether lessons can be learnt from it and applied more generally. As all the games are played over the Internet, but the teams physically assemble in their home locations, there are obvious applications to other remote-contact sports - particularly mental ones.
Baltimore turned out to be the winning team out of the eight in the first USCL season last year, coming good at the end of the season to shock New York in the semi-final and finishing with a flourish in the championship against Miami. Over the summer, the league expanded from eight teams to ten through new teams in Tennessee and Seattle. The USCL also gained sponsorship from PokerStars; not an obvious link, but many fine chess players are turning to poker as well these days. Expansion, sponsorship and increased awareness served to radically increase the strength of the league, which gained two GMs, one immediately becoming the league's strongest player, and doubled the number of >2500-rated players.
The other big news story of the summer was the move of Baltimore's top board (and league MVP for the first season) IM Pascal Charbonneau to New York over the summer, to play alongside his longtime girlfriend, the wonderfully-named IM Irina Krush. Three of the ten teams made it through the league phase of the second season winning more matches than they lost; there was a tremendously close scramble for the three remaining post-season spaces by six of the other seven teams. A rather neat post-season system has generated Western and Eastern champions, and accordingly the San Francisco Mechanics will take on the New York Knights for the second season championship a week on Wednesday.
Starting with the conclusion, I am very impressed with the job that IM Greg Shahade is performing as Commissioner. First off, you have to award a great deal of credit for dreaming up and taking on the enterprise in the first place. Secondly, add further credit for it not blowing up over two seasons' time; not only is there the natural tendency for the starting phase of an enterprise to be the trickiest part, he has been able to cope with all the events and unexpected happenings along the way. Third, continue to add credit to reward the fact that it's difficult to find anyone with anything bad to say about the USCL - and when they do, they're likely to be talking about the unfortunately similarly-named US Chess Live!, with which the United States Chess League has no connection.
All told, I think Greg deserves about 9/10 for the job he is doing. From this point onwards, I am going to start getting very critical, but only doing so in the knowledge that all the points I can raise are either very minor in the greater scheme of things, or issues where there are definitely arguments both in favour of taking the approach he has chosen to take and in favour of taking an alternative approach that I would propose. Lastly, in general, I think it's easier to be really picky about something of which you are broadly very strongly in favour; for instance, see the hatchet-job that the denizens of Bother's Bar did on Codex, despite the fact it's by far the most interesting - the only interesting? - new game show so far this year. (OK, the 888.com Poker Nations Cup was a fresh new take on poker.)
Particular plus points to celebrate are the achievement of sponsorship and the expansion of the league. (I fear that recent legal developments in the US will mean that online poker and online gambling companies will have rather less marketing budget to splash around, at least in the US, which might possibly restrict future sponsorship opportunities.) The move to add a Monday night game, distinct from the usual Wednesday night action, is fun, in the frequent sporting tradition, and the additional coverage on chess.fm is welcome. (I called it! Sadly EdgeTV, also referenced, appears to have gone cold, presumably terminally stone-cold.) In short, the expansion bodes well, and it's good to see that recruitment for further expansion is already underway. It's interesting to compare the call for 2007 expansion with the call for 2006 expansion - very little difference between them, but indicative of some of the little bugs being ironed out.
It should be noted that the shift in emphasis towards the teams promoting themselves and the league is a welcome and wise one; one might hope that widespread reknown among the chess fraternity is a step along the way towards wider name recognition elsewhere in the sporting world. Lastly, there's always the principle that "Nothing would be done at all, if a man waited till he could do it so well that no-one could find fault with it." Perfection is an unreasonable aspiration, even for a fan; merely doing as much good as possible and minimising the frequency and magnitude of mis-steps would put the USCL some way ahead of very many sports administrations.
On the minus side, it's very clear (perhaps clearer than the truth?) that it's a one-man operation, and a very hands-on operation at that. The web site looks dated and is more becoming of a local chess club rather than a national league. It would be wonderful to get a syndicated feed of the USCL news items as they are posted; my recommendation would be a nice content management system - possibly Plone or somesuch - where the architecture could be set up and someone less technical could add updates.
Greg seems to be very keen to be involved with all aspects of the league and knows that it's his show to do with whatever he will, at no notice. I can't point to anything specific, but I have half a recollection that he's got involved in comment threads on players' (and teams'!) blogs more closely than perhaps might be considered, well, in keeping with the tradition of the distant, all-knowing Commissioner. This is a fine line to tread; it's a reasonable complaint when a commissioner, or a commission, is too hands-off. However, any chess league will have more than its fair share of Type A personalities and sometimes the discussions get rather personal and heated. Displays of commissionerly power when required are the only way to make any league operate, but in this day and age it's very easy for any party in a dispute to put their side of the story forward. For instance, this apparent rule change at no notice. I don't know the full details of that case and it may well be the case that Greg's decision was the right one, but it all looks a little ad hoc. The MVP scoring system likewise looks a little, well, arbitrary.
Lastly, there is the aspect that Greg is the brother of WGM Jennifer Shahade, who manages the New York Knights. I'm not going to make any accusations of favouritism or malpractice and I haven't seen anybody make any, from which we can reasonably conclude that there has been no such improper conduct. However, I have this horrid feeling that this is something that's going to peek its ugly head up somewhere along the line. Not only must Greg do the right thing, he must be seen to be completely above reproach - and such family business, even if (as can confidently be expected) completely legitimate, will always seem a little... smelly.
Pluses and minuses identified, what of the future? Well, more of the same, only bigger and better, I suspect. Greg plans to add two more teams this year, and possibly every year in the future. Let's consider the practical implications of having 12 teams next year.
The USCL is currently split into two conferences, as is the American tradition. It's always worth investigating the reasons behind traditions, though. MLB and the NFL have two conferences each because they were formed from amalgamation of rival competitions; NHL and the NBA split their leagues geographically into two, so that teams play relatively local opposition more frequently, promoting local rivalries and cutting down on travel bills. However, on a global level, this multi-conference approach is anomalous.
It is to be noted that Major League Soccer, the USA's most prestigious association football competition, currently has two conferences - but according to DC United coach Bruce Arena this will soon stop being the case. This list of sports attendance figures tends to support the assertion that association football (soccer) is the world's premier team game; it is to be noted that only the US and Brazil, among even semi-major association football associations, do not have a unidivisional, national league as their top level. (Even the top two divisions in Russia are national, though Russia 2 looks like a huge trek.)
As every game is a home game for both teams in the USCL, and indeed any league where matches are conducted over the Internet, teams do not have travelling expenses for away (road) games, which would seem to be an argument against multi-conference organisation. However, this two-teams-at-home-simultaneously system does create a new problem; when the teams are in different time zones, the games take place at different times of day for the two teams. When a team operating on Pacific time plays a team operating on Eastern time, the match must be in the early evening for one team and the late evening for the other. This is going to be inconvenient for at least one team; in practice, this limits the length of the game to a lower duration than a game between two teams in the same time zone.
Accordingly, there does seem to me to be utility in encouraging games between teams in the same time zone. The current Eastern/Western split seems like an excellent way to do this, in theory, though it is to be noted that one of the teams in the Western division remains Miami, which is about as Atlantic as you get. (Six of the eight initial teams were in the Eastern time zone, with a seventh in the Central time zone.) If a one-hour time difference between teams is acceptable and a three-hour time difference is not, I would encourage the team to recruit two more teams from outside the Eastern time zone this year, so that the Eastern conference might be competed solely on Eastern time and the Western conference not at all on Eastern time. Naturally inter-conference play might cause three-hour time differences, but this is not avoidable.
Furthermore, a longer-term aspiration might be to have the Eastern conference contain teams only from the Eastern and Central time zones, and the Western conference contain teams only from the Mountain and Pacific time zones, to ensure no in-conference game ever has a time difference bigger than one hour. If the league is going to adopt a policy to prefer teams based in some time zones over others, this should be announced in advance, so teams in disadvantaged/over-represented timezones might know about extra difficulties their application might face in advance.
Further extensions to this could introduce a third or fourth conference to promote same-time-zone play, possibly by having separate conferences for single time zones - maybe even two conferences for the Eastern time zone. I would tend to discourage divisions within conferences, on the basis that having many small divisions can lead to the relative strengths of divisions causing an imbalancing effect with the best non-qualifier in a strong division being better than some qualifiers in a weak division, though possibly regional (sub-conference) titles for local rivalries that do not impact the conference-wide championship race would be an interesting addition. Trying to follow a model just because one particular sports league sets a trend would seem highly ill-advised, though.
It also turns out that two conferences of six teams would have the potentially desirable optional scheduling property that the (arbitrary) ten-week regular season might consist solely of intra-conference play (two matches against each of five opponents in the conference), leaving inter-conference play for the play-offs.
Another reflection is that expansion needn't necessarily be restricted to two teams per year. It's worth comparing the planned growth of the USCL with the growth of the 4NCL, the main British face-to-face chess league. I am not convinced that the 4NCL's situation is ideal - some of the teams have been short-lived, occasionally less than a season long - and many clubs have more than one team, which is also a source of contention that requires a decision. Nevertheless, this is an indication for the need for second team and minor league chess and I feel that the USCL would be wise to think about it, even if they decide not to take action on the matter. (The issue of second team and minor league competition in all forms of sport always presents itself; there is much good practice, but no established, globally-agreed best practice on this.)
It was mentioned above that MLB and the NFL owe their current composition to the merger of rival competitons. As the barriers to entry can never be terribly high, it is not inconcievable that a rival Internet chess competition might form some day. Suppose there are eight or ten applications for two expansion spots, leaving six or eight teams unsuccessful; if these unsuccessful applicants can find each other, it's possible that they might start a rival competition of their own. It also occurs to me that this rival competition might not impose the restraints on itself that the USCL has chosen; a putative rival World Chess League might not just feature teams from the US, but also add teams from Canada, possibly the London Night Owls, the Amsterdam Insomniacs, the Dubai Nocturnals, the Sydney Breakfast Club, the Hawaii We're Playing On Work Time and so forth. The market might well decide for itself here.
That said, that's partly in the realm of thought experiment. Rival competitions in sports leagues come about when there's money to be made from expanding the professional game, which seems unlikely for chess at the moment. More plausible sources of rival leagues would be sufficiently many people saying "Greg, we like the way you run your league, but we want to play at the weekend, not on a Wednesday" or "Greg, we like the way you run your league, but we want to play in the spring, not the autumn". Either could be plausible and might be a way for the USCL brand to extend itself without diluting the purity of the original USCL vision.
In conclusion, my overriding advice to Greg would be to prepare to spread the load - find a trusted team so that he might share the responsibilities once they become too much for one person and ensure that the operation of the league does not look as personal as it does at the moment. Delegation might be "the art of preparing to accept second best", but perhaps there might be specialists who could at least match Greg's performance in some aspects of competition organisation, though he has set a high standard to surpass.
The second half of this essay discusses how a hypothetical UK equivalent competition might be organised - how I might set up a counterpart competition in practice. In a way, it's nothing more than sports organisation fan fiction, but it doesn't pretend to be anything more. (Compare with the Global Twenty20 proposal for a different sort of cricket competition.) The generalities are also interesting because they could easily be applied to go, backgammon, Scrabble, team heads-up poker, shogi and so on, as well as all manner of other games potentially played over the Internet. The basic premise is that teams gather in a certain location to play whatever game it is, and that spectators can follow the game online or attend their local home venue to support the team in person. This is a crucial advance because it may permit team competitions in sports where budgets are too small for teams to conveniently travel to play each other.
It's worth thinking about what the ultimate aim of a new sport league is. This paper, although rather technical and somewhat above the level we need, identifies that sports leagues can have aims and that the aims are not always identical; for instance, "the league's objective may be either to maximize the demand for the sport or to maximize the teams' joint profits". It is to be noted that the former approach is more in keeping with European sports leagues and the latter with American ones. Indeed, it's not too far a stretch to assign those two behaviours to the 4NCL and the USCL respectively, what with their meritocratic club system and dynastic franchise system respectively. I would also ascribe a goal of "be regarded as the strongest league of its type" to many leagues; European leagues particularly frequently compete in inter-league competition to this end, plus all sorts of ego-boosts for league organisers and participants.
The financing of the league is also an interesting point. From the above paper, "Under demand maximization, a performance-based reward scheme (used by European sport leagues) may be optimal. Under joint profit maximization, full revenue sharing (used by many US leagues) is always optimal". It is relevant to note that the USCL openly offered sponsorship money evenly divided between the teams, whereas the 4NCL charges entry fees and redistributes some of this as prize money. It is also to be noted that the 4NCL operates a promotion and relegation scheme whereby teams in lower divisions are given a financial incentive to compete and fight through to higher divisions, where the prize money is higher. (Compare the 4NCL prize fund this year, specifically the relative entry fee receipts and prize fund disbursements from each division.)
It is also relevant to ponder what clubs would do with their prize money. Having significant financing, whether though prize money or sponsorship, is likely to attract strong players to participate. While this is desirable, it is to be questioned whether professionalism (i.e., players guaranteed payments to play) is (a) desirable or not and (b) can effectively be forbidden or not. Evolutionary patterns in amateur team sport tend to support professionalism as being inevitable, but deliberately lowering barriers to entry so that enthusiastic and organised amateur teams can compete in the same competition as the pros, possibly eventually on the same level as them. This is, at heart, a decision to be made on preference, with established sporting tradition supporting both options, but my sports-romantic European background prefers the idealistic possibility of no barrier to entry. (On the other hand, I can also see the attraction of creating the career of professional chess player.)
This naturally leads into the question of competitive balance. The (especially previously) moderately well-funded teams in the 4NCL have sometimes been able to attract extremely strong players, with the pinnacle being the final round 2004/05 championship decider between Wood Green and Guildford. Each league should make a decision about whether it is willing to accept such imports or require teams to be properly representative of their clubs, as happens in some competitions, though enforcing such a rule might well prove very difficult in practice.
There's also an approach to be taken by the league about the permissibility or otherwise of roster changes within the season - either transfers between teams or later additions to an existing roster. Again, there is no definitive right answer on how to implement this; arguably many, many wrong ones, though. As much as I enjoy football transfer gossip, I think the USCL's approach is not only a pure one but a wise one - and more flexible, it should be noted, than perhaps it first appears, because there have been few (no?) incidents requiring flexibility.
The USCL take a definite position on team strength by insisting that teams have an average (USCF, not FIDE) rating no higher than 2400, with a couple of twists. First, any female player in the team raises the permissible cap by 10 points (which is equivalent to the lady counting 40 points lower than she really is and the cap changing). Second, any GM over (some certain limit) counts as that limit. The limit started as 2600 and will be dropping by 10 points per season to 2550. To me, this suggests that the ideal USCL line-up (in three years' time!) would be three super-GMs, of unlimited strength but counted as 2550 for USCL purposes, and one player rated no higher than 1950 to make the average work. (If some or all of these can be ladies and raise the cap, permitting a stronger fourth team-mate, even better still.) This approach is controversial, but far from unknown elsewhere in sport; even in Britain, motorcycle speedway has a similar team-balancing mechanism.
This aspect of competitive balance strikes British eyes as very American - as American as the sporting draft. There is prior discussion about the desirability or otherwise of parity in sports leagues, and considerable argument on both sides; one interesting recent discussion was this post on the The Sports Economist blog, where former U.S. Soccer national team coach Bruce Arena discussed his preference for the more usual association football approach of little or no effort to achieve parity. My preference is against efforts towards parity (unlike my opinions about matters in real life!) but I recognise that this may well be mainly due to my sporting tradition upbringing and Greg's preference due to what he's familiar with.
The credit for playing female team members is further worthy of note. Promoting women's chess is desirable; promoting junior chess is desirable. (Promoting play among other groups who traditionally play little is surely also desirable; the Braille Chess Association, for instance, are very good at what they do.) It should be noted that the 4NCL insist that there is at least one female player (and at least one male player!) on each team, though this can sometimes be replaced by a junior male player in Division 3, and no restriction applies in Division 4. Most other chess leagues make no particular effort to reward diversity.
I also note that the likes of Judith Polgar and Magnus Carlsen show that female and junior players can compete on equal terms with men of any age, so any concessions are principally to encourage teams to have a diverse appearance in an approach that might attract more players to the game. My preference would be for a sports league that rewarded its teams financially less for excellent performance and more for widespread publicity, so promoting the league's aim of increased demand. Teams may choose to promote themselves through their excellent results, through their welcoming approach to diversity or quite possibly by both methods simultaneously. This might propagate a "the more successful get more successful still" problem, but whether this is truly problematic is a matter for discussion - see Bruce Arena above.
It is to be noted that both leagues take a similar attitude to lineup issues; in general, players must play in descending order of strength, though there is some flexibility to deviate from the rating list when players' ratings are close. While I see the wisdom in the approach, I tend to prefer the 4NCL's marginally greater flexibility and wonder whether there would usefully be scope for greater flexibility still. While we all want to see the strongest players face each other on Board One and the old saw about "why don't cricket teams put their bowlers to bat first then their batsmen in last?" has a point in that we would rather test chess than the ability to second-guess your opponent's line-up and manipulate your team's line-up accordingly, I think there are some interesting aspects of captaincy that being excessively inflexible here tends to deny. My preference would be to experiment with a slight loosening of 4NCL rules.
We shouldn't necessarily even regard the necessity for a particular number of players on the team as being sacrosanct, and uneven matches might better be resolved with handicap play than with obligatory forfeit; it might be interesting to see whether a single extremely strong GM could play a simultaneous display against a league team and beat them all at the same time - shades of Kasparov beating entire Olympiad teams in a simultaneous display and the like.
Let us briefly consider the possibility of cheating in an online chess league - of which, to my knowledge, there have been no accusations and so we can reasonably expect it not to have happened. My suggested workaround would be for each team to send one observer to the other team's physical location and ensure for themselves that there is no malpractice - no assistance from extra players or computer engines. An opposition observer could be useful in that a home-team-aligned observer might be reluctant to say anything that could damage the prospects of his own team. Of course, this necessitates travel expense and loses some of the benefits of Internet play, but surely better for one person to be required to travel than a whole team? I would also tend to suggest the burden of proof would be up to the opposition observer, presumably by catching the malpractice on camcorder tape, though this raises further expense and the theoretical potential for tampering with the evidence.
I would have thought that even if an official opposition observer were to be able to ensure that no inadmissible software was running and there was no obvious consultation, there would need to be some way of guarding against non-obvious consultation. I would have considered it useful to require customised playing accounts with much of the online chess functionality removed, so that players could be contacted by anyone who might wish to whisper accurate (or inaccurate!) computer move suggestions to them. Even then, it's hard to guard against a parallel situation to the "Toiletgate" allegations - perhaps a player might depart for the toilet, only to check his mobile phone for suggested moves that he might have been sent by SMS from an observer... This would tend to suggest to me that it would be wise not to raise the stakes too high.
Lastly, as I have spent so much time agitating in favour of expansion of the USCL, I would propose an aggressive expansion policy of any online sport league I were to run in the UK. I am a firm believer in the promotion-and-relegation system, which (if correctly handled) relatively quickly permits teams to find the most suitable level of opposition for themselves and yet provides an appropriate and exciting competition at each level, with new entrants able to feel they are within touching distance of making their dream of facing the best in competition come true. (FA Cup-style unseeded random cup competitions are also inherently good and so to be encouraged, should the schedule permit.)
As I get to play fantasy league organiser in this post, I would agitate for a relatively flat pyramid from the very start - one national top division, two half-national second divisions, four quarter-national third divisions, eight increasingly regional fourth divisions and so forth. (This, of course, relies on very convenient numbers. Hypothetically, as much as 1-2-4-8-16 would be ideal, 1-2-4-8-8 would be an acceptable intermediate step, and I would prefer 1-2-4-8-8 with larger fifth divisions than 1-2-4-8-12 with divisions of equal size, and so forth.) Parallel divisions on the same level would be equal in stature; as they would all be organised by the same competition, there would be no problem with moving teams from parallel division to parallel division within a level, should the geography suit better all around.
The plan would be to reach deeply within the talent pool, for a large but feasibly conquerable competition that could involve as many people as possible. (With appropriate organisation, naturally.) Regionalism works well at a lower level, as much as I discourage it at the top level, given that all the teams at a lower level know they are steps removed from a single national championship. It is to be noted that many weaker Scottish teams prefer to play in the Scottish National Chess League rather than the 4NCL, which can probably be assessed as a failure of the 4NCL to cater to Scottish clubs. Perhaps the 4NCL might consider several parallel regional lower divisions to make the travelling burden for distant clubs less.
The fun part is getting to pick the league names! Naturally, accurate league descriptions would depend upon the geographical make-up of their constituent teams, and it seems wise not to get too specific about the areas. (After all, a "Scottish" league might well have too many or too few Scottish teams, yet a league whose name has a clear hint of Scotland without being too specific would work well.) I would favour splitting a second division into a Provincial conference and a Capital conference, roughly NW/SE, a third division into North, East, West and South conferences and a collection of eight suitable names for a fourth division might be Alban, Northumbrian, Midland, South-East, Greater London, South-West, Cambrian and North-West. The keys to me appear to be not to repeat a name at more than one level and always to be slightly vague. Should a league get to a fifth level and need 16 new conference names, that would be a delight for the future!
A brave new world of organised online competition beckons - and, just think, all it would take for me to make it happen would have been to hit the EuroMillions...
Current Mood: thoughtful
|Date:||November 21st, 2006 06:43 am (UTC)|| |
hello from the USCL Commissioner
Hi, I have a lot of specific reasons for some of the rules that are implemented, which aren't addressed in your posting. I am planning on writing an FAQ on my website going into more details about some of the league's decisions, the format etc etc.
If you ever wanted to email me a list of questions or something like that, I'd be glad to respond to them. I don't know if that makes sense but whenever I see discussion about certain rules or parts of the USCL, in which the discussion fails to notice a few points, I get antsy :)
Anyway the project is still in it's growing stages. I don't believe we have the capabilities for rapid and grand expansion just yet, hence we are going slowly. There are going to be a few cool additions to the website in the next few months, although some of them are already set, I plan to add one or two new things every month or so to keep things fresh during the long offseason.
Also I will give you a little sneak preview of the 2007 format.
2 divisions, 6 teams per division. You play 3 teams in your division twice, the other 2 teams once. Also there will be 2 interdivision games. The schedule will be based on previous year's performance, with the teams that faired better getting an easier schedule (or at least easier in the sense that they will be playing more teams that didn't fare as well). There is an exact formula that has already been decided to create the schedule.
|Date:||August 16th, 2008 02:06 am (UTC)|| |
Your blog is interesting!
Keep up the good work!