A Platform Tennis Index (PTI) is a performance rating developed by the APTA in partnership with Paddlescores, LLC. PTI is an objective data driven measure of rated match performance. Using this players, captains, teaching professionals, league administrators and tournament directors are able to organize competitive play amongst players of similar performance.
While a player’s PTI reflects his/her ability, it is more precise to describe it as a measure of rated match performance. Players who routinely defeat the same opponents 6-0, 6-0 will have a stronger (lower) rating than those who defeat them 7-5, 7-5.
This could be because the 6-0, 6-0 winners were simply so much stronger than the 7-5, 7-5 winners. Or because the 7-5, 7-5 winners decided to keep the match “friendly.” This is consistent with the goal of using PTI to organize “good matches” vs. “not so good matches.”
Regardless of how initial PTI ratings are set for players, the APTA and League commissioners will review player’s match results to determine if their performance was consistent with the initial start ratings and make modifications if the early matches show it was off.
More information: Initial Start Ratings - Defaults and Overrides
PTI ratings are a dynamic system and are updated frequently to incorporate new and corrected match results and player ratings.
Primarily, PTI ratings are adjusted after each match. The PTI algorithm uses the match type, match results (winner and game/set score), and the ratings difference between the 2 teams to calculate the adjustment.
For example: Alice (20 PTI) and Bob (28 PTI) play a match against Charlotte (24 PTI) and Doug (30 PTI). Alice & Bob have a team PTI of 48 vs Charlotte & Doug (54). Alice & Bob have a lower team rating and are favored to win the match.
If Alice & Bob win, this is the expected outcome, and will likely see a small decrease in rating to reflect the margin of victory. Charlotte & Doug will see a small increase of the same amount.
On the other hand, if Charlotte & Doug win, this is an upset and will be reflected by larger adjustments to reflect the unexpected outcome.
It would be reasonable to say that a PTI is changed through unexpected results (upsets) and fine tuned during expected results (when the favored team wins).
Starting with the 2019-2020 season, the APTA is incorporating a Confidence Factor into the PTI algorithm. This innovation provides another dimension to further improve the adjustment of a player’s PTI.
Confidence Factor ranges from 0% to 100%, and increases with each match played. A player reaches full confidence (100%) at eight rated matches played.
For a posted match result, players with lower Confidence Factors will have less of an impact on the PTI of players with higher Confidence Factors. At the same time, the ratings of lower Confidence Factor players will adjust faster than others. This allows players to reach an appropriate PTI more quickly.
More information: Ratings Confidence
There is no way to change your “current” PTI. Only your initial start rating or the confidence factor on your current PTI can be changed. Once/if one of those are changed, the algo will go back (overnight) and recalculate every match you’ve ever played to produce your new PTI.
To have either your initial start rating or your confidence factor changed, you need to petition your league’s PTI Committee. This is typically done through your team’s captain with an explanation for why an adjustment makes sense.
While the rating system is still new to some leagues, we are allowing League Commissioners to change initial start ratings for players in their respective leagues. Eventually, we’ll only allow initial start ratings to be altered by the APTA PTI Committee. A player or league commissioner will be able to petition the Committee in writing to propose a more accurate initial Index.
The Confidence Factor may be altered at the league level by a League Commissioner or League PTI Committee. Reasons to alter a player’s Confidence Factor include significant injuries, long absences (multiple months or years) from the game, a strong tennis player who is “figuring out” paddle, and rapidly improving skill.
A golf handicap measures a golfer's “potential” because it throws out your worst scores and takes an average of your best scores. Your PTI doesn’t measure your potential because it captures all your matches. It’s a measure of your actual performance and not your potential.
The APTA has formed a PTI Committee currently made up of 9 people from 7 different leagues across the country. This committee monitors how the algo is working and reviews suggestions for tweaking the algo during the off-season.
There are no standards currently in place for how a league could or should use the PTI system. Each league decides for itself whether and how to use PTI ratings. Some of the more common uses for PTI include:
While it’s a rare occurrence, it is possible to have your PTI get worse even if you win a match. The scenario where this happens is when a heavily favored team barely beats its opponent. While the favored team gets the winning bonus from the algorithm, the percentage of games won can miss expectations by enough that the winning team’s PTI gets a little worse. The move is almost always only a few tenths of a point.
This is the corollary to the previous question. The algorithm rewards winning first and percentage of games won second. Had the underdog team pulled off the upset and gotten the win, the PTI adjustment would have been large - well over a point, possibly close to two points.
There are a couple instances where the algorithm will skip a match. The most common is when two teams playing each other have a combined PTI spread of 21 points or higher. That match won’t count towards PTI UNLESS the underdog team actually wins. When the combined ratings spread between two teams is between 16-21 points, the algo starts to decrease the weighting of the match (unless again the underdog teams wins). This is to prevent heavy favorites from being forced to blow out weaker opponents.
The other scenario where a match won’t count towards your PTI is if its marked as “Default."
If one full set has been played, the match will count towards your PTI. If less than a full set has been played, the match should be marked as “default” so it will be skipped. Leagues and tournament directors may make exceptions to this rule if they feel its appropriate (example: a team is ahead 6-0, 5-0, 40-love…and someone from the team ahead gets injured; a league or TD may determine not to award the win (for PTI purposes) to the uninjured team that was on the cusp of losing).
It shouldn’t. A weaker partner comes with a higher (worse) PTI, which means the expectations for the outcome will be reflected in that higher PTI. It doesn’t matter whether your partner is better or worse than you…what matters is whether they can play to their PTI. You’re better off playing with a player rated a 50 who plays like a 45 vs. a player rated a 5 who plays like a 10.
Main draw tournament matches count 1.5x a league or ladder match. Back draw tournament matches count the same as a league or ladder match.
Any league, ladder, or tournament match entered on a Paddlescores site will count. Casual matches entered on a Paddlescores will also count but only 50% of a league match. Any APTA tournament that uses Live Scoring where the tournament director has entered scores properly will be imported into the PTI database (and count toward your PTI).
If any players in a match are new to the PTI system, their PTI will have a confidence factor (CF) less than 100%. Players with CFs less than 100% will adjust more quickly than players with CFs of 100%. The lower the CF, the greater the adjustment. Your PTI will adjust the exact same as your partner’s if/when you have the same CF. Everyone’s CF reaches 100% once you’ve played seven matches that have counted toward your PTI.
The core PTI algorithm is a follows:
This core algorithm is an implementation of the ELO Rating System used widely in competitive sports and activities.
The PTI system adds additional adjustment multipliers based on:
More Information : Expected and Actual : What do they mean?