stalkor wrote:
The idea of the league is not to provide a full month of play but to provide a competition. Moost of the alpha players are very active, for example joenosai tries to play all games each month and he does most of his games early in the month. 3 player are not active this month, one of them having quit last week and this is why we are at a lower percentage then usual.
Most alpha would rather see a decrease in games to increase quality but they understand the fact that we need it this way and happily play anyway.
So in the case of alpha the inactivity in the last part of the month is either a too different schedule or a few quitting players.
With regard to the last point, here is the same table for November. There was quite a different pattern of play with much more of the activity for all classes concentrated at the beginning of the month. Nevertheless we can see the same type of decline in activity. The explanation that this is due to the unexpected activities of a few individuals is not very convincing IMHO.
Having said that, I do have to admit that looking back over the results for more months (prior to my "hands on" experience in the ASR) the activity level in alpha has varied considerably. Since it was reorganized as an 18-person class with two games each last February, the level of games completed has ranged from 30% (April and August) to 65% (July). Alpha has not always been the most active class.
The idea that alpha is about competition may represent the original desire when the ASR was founded, but it does not fit the current structure and activity in the league very well. What is the nature of the competition? The league scoring is evenly split between activity (games played) and winning. As a result, all the classes mix players of different strengths and activity levels. Alpha is no exception. What would be some signs that it were providing competition? Surely the primary one would be that the results of the players were closely comparable, right? Yet, they are not. A glance at the results table in my earlier post easily show that alpha members are rather easily divided into those who win most of their games and those who do not. Alternatively we might expect that if alpha is competitive a more detailed look at the results would show that in cases where the same players played two games against each other, the results tended to be split with one win and one loss each. If the results are random, or you might say completely competitive, we would expect the number of split results (red-green or green-red) in the table to be equal to the number of double wins (green-green) plus the number of double losses (red-red). That is not actually the case. On the 45 cases in the December table (so far) where two players have played a pair of games with each other, 37 times the same player won both games whereas the players split their games only 8 times. To put it another way, the repeated win (loss) paired games are occurring about 4.5 times more often than the split pairs. Looking back to February again, this ratio has varied from a high of 11.5 (April) to a low of 1.7 (August). The average of the monthly figures (February through November) is 4.4. It has never approached 1. So if alpha is about competition, it is not clear what the nature of the competition is.
I do not think the present league lends itself to competition - here I assume that the competition we are talking about is games with an equal likelihood of being won by either side. If the ASR wants to sponsor competitive activities, those activities will have to be in addition to the league and structured differently. They will have to bring together players of similar strength. That will necessarily be a departure from the league idea of people of different strengths playing seriously together and reviewing the games together afterward.
The present league scoring rewards both activity and winning. As a result, alpha consists of a range of people who either win a lot and are at least reasonably active or who play a lot and do not necessarily win so much. Since in the league in general each class is smaller than the one below it (and therefore has a lower number of possible games), we can consider that there is a gradual indirect bias against those who are active compared to those who win as we climb the league structure. This has not prevented people with lower winning percentages from winning alpha (for example see October in the results archive). It would be quite interesting to see the evidence for claiming that the members of alpha would prefer to play fewer games if that increased "quality". I was not involved with the ASR a year ago, but if I understand what happened, this is what was attempted with the "superclass". That did not turn out to be popular and was dropped due to inactivity (only 4 games played in February), right? I would definitely like to see that idea compared to one of increasing the number of games available in the current format. Personally I would wager that with the current class composition, the alpha members would prefer more opponents than the opposite. As mentioned above, more competition would be very interesting, but I believe it requires a set of activities that are additional to the league.
_________________
Dave Sigaty"Short-lived are both the praiser and the praised, and rememberer and the remembered..."
- Marcus Aurelius; Meditations, VIII 21