The candidate leagues in this case are the 55 & Over Adult leagues with 6.0, 7.0, 8.0, and 9.0 flights where players up to a full two levels apart can play together. For example, in 8.0 the playing combinations can be 3.5/4.5, 4.0/4.0, and 4.5/3.5. The fundamental question is, when different pairings face each other, which wins more often?
Note for this analysis I'm only considering "at-level" pairings and not those where say a 3.5 and 4.0 play together as a "7.5" pair in an 8.0 flight.
I've done this before for Mixed, both 18 & Over and 40 & Over and in nearly every case, unbalanced pairs win more often than balanced pairs, generally between 52% and 60% depending on the exact scenario. Mixed also has different genders to add to the permutations but the strongest tends to be unbalanced with the male the higher level player where that pair wins around 60% of the time.
What I did was to take a look at Adult 55 & Over matches from 2022 and broke it out by both gender and level.
For the women:
- 6.0 - 3.5/2.5 pairs beat 3.0/3.0 pairs 51% of the time
- 7.0 - 4.0/3.0 pairs beat 3.5/3.5 pairs 55% of the time
- 8.0 - 4.5/3.5 pairs beat 4.0/4.0 pairs 57% of the time
- 9.0 - 5.0/4.0 pairs beat 4.5/4.5 pairs 66% of the time
We see unbalanced pairs have the advantage, very slight at 6.0 but growing as the level goes up.
For the men:
- 6.0 - 3.5/2.5 pairs beat 3.0/3.0 pairs 57% of the time
- 7.0 - 4.0/3.0 pairs beat 3.5/3.5 pairs 56% of the time
- 8.0 - 4.5/3.5 pairs beat 4.0/4.0 pairs 56% of the time
- 9.0 - 5.0/4.0 pairs beat 4.5/4.5 pairs 57% of the time
The men are very consistent at 56-57% of the time the unbalanced pair wins.
This is pretty consistent with the Mixed results, unbalanced pairs win more often. They theory is that the unbalanced pairs generally have the strongest player and if they employ a strategy to minimize the lower rated partner's weaknesses, the stronger player rises to the top. At 55-57% for most scenarios it is obviously not a given that unbalanced pairs will win, but the trend is pretty clear.
What do you think? Is this consistent with what you see in your leagues?
Update: I did a subsequent deeper dive looking at actual vs expected results here.
I don't play this division, but probably makes sense. But, I think it moreso comes down to just finding the top players at each level or closeish to this but I guess not entirely. And it's much more easier to find such players as the levels go lower, such as easier to find a top 4.0 than a top 4.5, a top 3.5 than a top 4.0, etc. So, I'd think a 4.5/3.5 combo where the 3.5 is top of level and even if the 4.5 is just average or slightly below average for 4.5 should normally win against 2 4.0s, unless both 4.0s are near or at top of their level. The 3.5 probably isn't too far off in ability from the 4.0s normally.
ReplyDeleteRegarding mixed, I see the 4.5M/3.5F combo in 8.0 normally perform the best. But, I see the 4.5F/3.5M combo perform the worst. But, there's so many more 4.0/4.0 combos, which most of these are average or below average courts, so maybe the 4.5F/3.5M combo does perform better overall. I wonder how it looks now?
Obviously not entirely, but what I see it coming down to primarily is whoever has the best player on the court, for any kind of doubles. Whoever that side is, will win 85-90%+ of the time.
Good point about the unbalanced pairs often simply having highly rated players at the lower level. What I need to do is go deeper to see if my ratings actually say the unbalanced pairs are favored to win.
DeleteYes, I do. Much like in Mixed, I often see a tendency to have very strong players (either S or C rated) who play with the higher partner and do really well. Key reason why I believe at-level play is better competition all the way around, not this 2-level difference. I do understand the argument of trouble fielding teams in Sections/Areas where the tennis population is not as large, but, from strictly a competition POW, at level play would be better.
ReplyDeleteIf someone is S rated playing mixed, I think that's a different scenario and obscures the results more. Not all S players, but obviously a lot of S players are often ringers or close to it.
DeleteKevin's analysis from 8 years for 7.0/8.0/9.0 shows the uneven pairings are winning 51-62% of the time. The average is probably 55-57% range, which is barely above average. This isn't extreme at all. Maybe if it was 70-75%+, then I could see it being a problem, but it isn't currently. For mixed(not 55+) the uneven pairings shouldn't be a problem with the woman being the higher rated player either as she would still unlikely be the best player on the court and often not even between her partner and herself.
The main problem I'm guessing that the USTA would humor only at-level players is that enough players have complained about losing and losing badly too often. I understand that nobody wants to lose and/or lose badly, but there needs to be more than this. Kevin's analysis doesn't show it being extreme either. These players are most likely near the bottom of their levels, say 4.0s playing 8.0. They're not winning much regardless if uneven pairings are allowed to play or not. But, the pool of eligible players for each division would be lowered dramatically.