Purpose. This study introduces the concept of Individual Winning Consistency (IWC) as an Opponent Quality metric in Freestyle Wrestling. The research aims to address two primary questions: a) Can a higher IWC for a contender, calculated prior to a match-up, reliably predict the match result? b) Can the IWC effectively predict Medal Performance? Methods. Athlete Place and Match results of the Wrestling Senior World Championships 2022 were collected, with 176 Women`s (WW) and 258 Men`s Freestyle (FS) participants. Final athlete rankings were compiled for those participating in at least one international tournament between September 10th, 2020, and September 9th, 2022. The quantification of IWC involved recording victories for each individual and subsequently calculating the median victories per tournament. For instance, an undefeated wrestler winning 10 tournaments, each with 4 victories, would hold an IWC of 4.0. Match pairings were meticulously chosen based on varying IWC values, enabling computation of the associated metric "Winning Consistency Advantage" (WCA) by subtracting the lower IWC from the higher. For instance, in a match-up between athletes with IWC values of 3.5 and 1.0, the athlete with higher IWC would have a WCA of 2.5. A total of 151 WW and 236 FS matches formed corresponding datasets, each split into training (70% of the data) and testing (30%) subsets, for CHAID classification tree analysis to evaluate outcome prediction. The wrestler's outcome ("won" or "lost") when in advantage was the target, with three WCA value ranges categorized as "Big" (WCA >=2.5), "Moderate" (WCA of 1.5 to 2.0), and "Small" (WCA of 0.5 to 1.0) as predicting variables. Medal Performance prediction utilized IWC of all the participants with international participation in the 2 years prior time span as predicting variable and final place as target variable, ranging from 1 (1st place) to 24 and 34 (24th place in WW and 34th place in FS respectively), treated by regression CHAID. Accuracy %, p-value, and Root Mean Square Error (RMSE) were calculated for model evaluation. Results. The accuracy of Match Outcome prediction achieved by WCA was 82% for the WW testing dataset and 72% for the FS testing dataset (p < 0.05). The calculated probability of winning a match was greater than 75% with a "Moderate" advantage, and greater than 85% with a "Big" advantage. Medal performance prediction reached RMSE lower than 0.4, indicating IWC >= 3.5 as predictor of 1st to 3rd place in WW (p < 0.05) and 1st to 8th place in FS (p = 0.00). Conclusion. Evaluation of classification and regression models demonstrates that IWC and WCA, computed from international performance data spanning two years prior, predict podium performance and match outcomes within the context of Senior World results. These indicators hold promise for high-performance Olympic Wrestling programs, contributing to performance prediction, athlete selection, resource allocation, and opponent scouting.
© Copyright 2023 International Journal of Wrestling Science. Alle Rechte vorbehalten.