This is a second post related to investigating the ambiguity of conclusions related to hitting efficiency. A Look into the Wentworth Leopards matches in Week #2 of the volleyball season propelling them to the AVCA #1 rank.
Most agree hitting efficiency in a volleyball match is a strong predictor for who wins. About 95% of teams with the higher HIT% win the match. Through last evening on January 24th, 149 of the 158 winners earned a higher attack efficiency (94.3%). For comparison, the team scoring the most points wins a match 96% of the time, only 1% more likely to win scoring more points than having the better HIT%. (In many sports the team with the most points will always win, but in volleyball, being the first to win 3 sets doesn’t guarantee the loser of a match won’t score more points. A losing team to score more points frequently lost a 3-2 match, and in rare cases a 3-1 loser might score more points. i.e. Winning one set by a higher margin than the sum of the 3 which were lost is very uncommon.)
In stats, a 5% threshold is often considered rare enough to warrant a reason, other than chance alone, for why something happened. The same as the chance for a volleyball team to win a match with a lesser attack efficiency than their opponent. So when this 5% event happens two consecutive times (p = .05 x .05 = .0025), it leaves the thought, “Which is more plausible? I just saw something with less than a 1 in 400 chance to randomly happen and it did, or I just saw something this rare because of some yet to be explained reason other than chance.” This thinking leads to high confidence conclusions for something more than meets the eye. Unfortunately, it offers no road map telling how to search for that “something more.” What does any of this have to do with week #2 in D3 men’s volleyball?
Wentworth, the new #1 team in D3 in the AVCA Coaches Poll, won two matches last weekend, each time with a lesser HIT% than their opponent. First, a 3-0 win at MIT, being outhit by the Engineers .282 to .260, a condition itself so rare it happens in well less than 1% of all volleyball matches. i.e. Sweeping while being out-hit. Wentworth then proceeded to defeat last year’s National Champion, and the previous consensus #1 ranked team, Stevens, by a score of 3-2, being out-hit again .224 to .209. A phenomena which happened in just 9 of 158 matches this season happened in back-to-back matches for Wentworth.
Hitting efficiency measures the effectiveness for successfully terminating when possessing the ball. To be less effective than another team in executing this skill, yet still win 6 of 8 sets played might be explained by earning enough points before possession to compensate the lack of those earned after. i.e. The serving game. I notice Wentworth served 16 aces/27 errors to their opponents’ 6 aces/33 errors on Saturday. This begins to shed some light on how a seemingly 1 in a 1000 event happened.
As it pertains only to “knocking off” the previous #1 and defending National Champion, Stevens, Wentworth doubled their aces while perpetrating just 20% more errors from the stripe as Stevens did. (A: 10-5 & E: 18 to 15) I reached out to Coach Giglio to see if their serving put Stevens out-of-system more than Stevens did to his team. He agreed the “The service pressure has been a constant for (us in) winning big matches.” Later, he replied again to state he thought passing was nearly equal, stating his preferred metric, “We were both at about a 2.2 for the entire match.” Knowing 3 of the last 5 serves Wentworth made resulted in Stevens passes to qualify as such, the expectation would have been for Wentworth to be about 25% more successful in forcing out-of-system passes in the match, maybe along the lines of 20 to 16. (+4 in that category) If only close to the truth, then the entire 4 point overall scoring margin links back to Wentworth’s superior serving.
Stevens ability to terminate, particularly on first ball side-outs, is such that any opponent will not win the match, even should they superiorly serve the ball. For example, doubling an ace plus out of system pass rate a service game produces won’t often be enough to compensate, unless the opponent approaches Stevens rate of first ball side-outs. North Central found out in last year’s championship this very thing as they produced aces or out-of-system passes in 1 of every 4 serves they made compared to Stevens producing just 1 in every 8. NC served superior to Stevens in the 2023 Championship, and it still was not enough to tilt the final outcome their way.
Explaining this series of events, likely not to happen again the rest of this season, as seen through a “service lens” gives meaning as to why it happened. In other words, the more points derived via serving, the lesser the points required to win through live-ball play, thereby allowing for Wentworth on two consecutive occasions to produce a lesser kill rate, but still be enough to win the matches. An equally interesting investigation for another day might be the trend in how each team’s hitting error rate changed throughout the match allowing Wentworth to come back against Stevens. This noticed when realizing more than half of all Stevens hitting errors took place in the last two sets, one of them only to 15. i.e. 52% of all errors in the last 35% of all points played.
The blue-print is now available to defeat Stevens, thanks to Wentworth. The margin any opponent serves better than Stevens, both acing and putting them out of system, must compensate for the margin Stevens first ball side-out rate betters the opponent. One way this happens is treating service like a poker player treats his chips in an “all-in moment” so the reward has a chance to justify the risk. Otherwise, to continue the poker reference, your chip stack will simply be bled away until you are asked to leave the table. Ironically, the blue-print for defeating Wentworth was also shown, even in two consecutive victories propelling them to the top spot in the AVCA rankings this week. If your team has the capacity to terminate with them, as both MIT and Stevens showed they could do this weekend, then do whatever it takes to not lose the service game. ERRORING 2 TO 1 SIMPLY WON’T GET IT DONE!
And to all those aspiring to execute a fine tuned dive into either plan indicated by these fairly dubious blue-print designs , “Good Luck with that!”