The general observation has been that unbalanced pairs, e.g. a 4.5 and 3.5 in an 8.0 flight, tend to win more often than not, and this deeper dive will take a look at if this is actually expected or not based on the current ratings of the players.
The deeper dive on 55 & Over revealed that while the unbalanced pairs do win more often, they are always expected to, and often fall short of the expected winning percentage. This may indicate a selection bias of sorts, where it is possible that only stronger players at their level opt to play as an unbalanced pair, thus skewing the stats in their favor, and there is not some inherent advantage to unbalanced pairs.
So on to taking a closer look at Mixed. Specifically, I narrowed the matches analyzed to only those where I had an established rating for all 4 players and then looked at not just how often unbalanced pairs won, but how often they were expected to win. The results were certainly interesting, and a bit different from what we saw for 55 & Over.
Note, the actual winning percentages are slightly different from the earlier post because some matches where the players didn't all have ratings have been filtered out.
For each level/gender below, I'm using 2022 18 & Over Mixed league data again and showing the percentage of the time the unbalanced pair won, followed by the percentage of the time the unbalanced pair had the higher combined ratings. When the actual is higher than the expected, the unbalanced is overachieving, and when the actual is lower than the expected, they are underachieving.
The ratings used for this analysis are my Mixed Exclusive Estimated Dynamic NTRP Ratings I use for doing the various individual, team, and flight reports I offer. The pairings are showing the male then female levels.
Starting with 6.0:
- 3.5/2.5 vs 3.0/3.0 - 52.0% actual, 57.1% expected - underachieved
- 3.5/2.5 vs 2.5/3.5 - 53.7% actual, 41.1% expected - overachieved big
- 2.5/3.5 vs 3.0/3.0 - 48.0% actual, 59.3% expected - underachieved big
The 6.0 level can be tricky with a lot of new/unknown players at times, but we see the pairs with the higher rated male player win more often, and this is especially true when it is a 3.5 male vs a 2.5 male and the 3.5 male really overachieves vs expected.
But like I said, 6.0 is tricky, so on to 7.0 where players are a little less volatile:
- 4.0/3.0 vs 3.5/3.5 - 58.7% actual, 56.4% expected - overachieved a bit
- 4.0/3.0 vs 3.0/4.0 - 56.3% actual, 51.7% expected - overachieved
- 3.0/4.0 vs 3.5/3.5 - 49.7% actual, 53.3% expected - underachieved
We again see the pair with the higher rated male wins more often, and we also see some over achievement for the 4.0/3.0 pairs.
Next, 8.0:
- 4.5/3.5 vs 4.0/4.0 - 57.8% actual, 54.9% expected - overachieved
- 4.5/3.5 vs 3.5/4.5 - 52.4% actual, 48.9% expected - overachieved
- 3.5/4.5 vs 4.0/4.0 - 54.2% actual, 53.2% expected - overachieved a bit
The trend of unbalanced pairs overachieving continues by a decent 3-4% margin for the 4.5/3.5 pairs.
Next, 9.0:
- 5.0/4.0 vs 4.5/4.5 - 61.6% actual, 53.6% expected - overachieved big
- 5.0/4.0 vs 4.0/5.0 - 59.1% actual, 55.0% expected - overachieved
- 4.0/5.0 vs 4.5/4.5 - 56.6% actual, 54.6% expected - overachieved a bit
The over achievement for unbalanced pairs is seen here too.
The counts for 10.0 are so low there is no point to doing the analysis here.
Given we see some to significant over achievement for the unbalanced pairs, especially at the higher levels, it may be that there is an inherent advantage in Mixed. This is perhaps because at higher levels, players are able to employ strategies maximizing the difference in abilities taking advantage of having the stronger player.
It could also be that for the lower level player of unbalanced pairs, only the top players at that level self-select to play, and that gives an inherent ratings advantage over two "average" players on a balanced pair.
What do you think?
When you say you're using the Mixed Exclusive Ratings for this analysis, do you mean are you using the ratings from these players from only their mixed matches even if they have C ratings? Players' ratings from adult league matches will be different from their mixed matches. Players are often better in mixed than adult matches or vice versa.
ReplyDeleteLooking at this overall, the unbalanced pairs are expected to win more often which makes sense. Just speaking from 8.0 experience, almost no 3.5F would ever be below average rated who'll play 8.0, and usually they're high, very high, or even over the 3.5/4.0 threshold. In contrast, the 4.5M or 4.5F playing 8.0 will sometimes be below average rated as a 4.5, though I'd expect a lot of the 4.5s will be rated at least average.
It makes sense that when the unbalanced pair has the higher rated male, that pairing wins most of the time for each division, and actually overachieving for each except 6.0, but 6.0 can be trickier to analyze. The side with the higher rated male will have the best player on court almost always, so naturally the side with the best player should win usually.