While I don't agree that AllegSkill is or has worked, I'll presume that it's what the community is tolerating.
I'd like to say that a YTD (Year To Date) rank reset is a good thing, but that a LTD (Lifetime To Date) is important.
As we see since the last rank reset, is that some vets aren't getting ranks fast enough, this is yes part of their choosing the wrong teams or having unexpected results. However the main problem is that after 50 games some vets are still ranks under (5), this is a huge issue in my humblest opinion (my rank is (10) currently and going up so I'm not talking about me).
Some solutions;
1.
I have one huge issue with AllegSkill (as do many others), is that once it stabilizes at what it 'thinks' a players rank should be, it rarely if ever moves. Sigma needs to not stabilize around 1 but instead a higher value, this should be done by increasing Gamma as has been originally suggested by the designers of AllegSkill if the very problem discussed here happens.
2.
Many people have noticed since we adopted AllegSkill that it doesn't reflect increases in players skill, while taking measures as stated in my first point will help, it is important to maintain a YTD regular annual rank reset (in part), and also maintain a
LTD record. The reason is outlined in my third point below.
3.
Sliding record, which basically means that your YTD will show your rank for the year but the LTD will be an important part as well. While the exact mechanics can vary what I suggest is this after an annual rank reset.
A) Your YTD rank is basd 50% on your choices for the year and 50% on your historical rank.
Ai) You start the year with 100% of your LTD rank, each game you play that changes more equally to your YTD performance. If you are performing worse than you'll see your rank drop, if you perform as good or better than before your rank will stabilize and then start showing improvement (each of the first 50 games decreases the LTD portion by 1% and increases the YTD by 1% until they are 50%/50%):
Code: Select all
GAME LTD YTD
1 99% 1%
10 90% 10%
50 50% 50%
At the end of the year your LTD performance is adjusted with your last YTD performance.If you played 1000 games with rank 10 for LTD and then played 1000 games at rank 12 for YTD, your LTD rank averages now to 11 for example. Generally speaking it will be difficult to increase your LTD by much as more and more games get slotted into LTD and you'll end up playing less games YTD. However you will see improvment to your LTD if you perform better.
Edit; Option, For more clearly shown performance improvements; before sliding in the YTD games into LTD, remove ½ of the number of YTD games from the LTD games, this will show a more noticeable change from year to year.
Suggest we also fully adopt the 50 point rank scale as originally intended by the authors of AllegSkill.
These are only suggestions, and please take them as well meaning as they are meant to be.
