What if the new 4 court format was used for 40 & Over in 2019? There would be unbroken 2-2 ties!

For 2020, the USTA is moving to a new four court format for 40 & Over USTA League play at the National level.  Specifically, a team match will have just 1 singles match and 3 doubles matches.  This is a change from the current five court format with 2 singles and 3 doubles matches.

Note that while this is the format that will be used at Nationals, sections have the freedom to use alternate formats for Sectionals, Districts/States, and local league play.  Some areas already have used 1 singles and 3 doubles, or 1 singles and 4 doubles formats and similarly, in 2020 an area could choose to stick with 2 singles and 3 doubles, but my guess is that a lot of areas will adopt the new format to be consistent with Nationals.

In what I wrote summarizing the change, I raised a number of issues, a significant one being that with four courts, a team match tying 2-2 is very likely to happen quite frequently, and then the team match winner is determined by tie-breakers, these being the fewest sets lost, fewest games lost, and then best percentage of games won.

A natural thought I had, and several readers asked me to do it as well (thanks for the comments and feedback!), was to look at past matches in 40 & Over to see what the potential effects of this rule change are.  If you are into numbers and statistics, please read on.

Default Rates in Singles

First, I looked at how often 2 singles is defaulted in both 40 & Over and 18 & Over play, the theory being if there was a problem fielding singles players in 40 & Over, there would be a noticeably higher rate of defaults on this court than in 18 & Over.

This turned out to not be the case as while 40 & Over had a slightly higher percentage of defaults, 6.6% of courts vs 6.4% for 18 & Over, 18 & Over actually had a higher rate of defaults on 2 singles, 38% to 36% for 40 & Over.  So the data does not seem support the idea that teams in 40 & Over are having to default 2 singles at an abnormally high rate due to lack of players willing to play.

While the lack of defaults could be due to a doubles player sucking it up and playing singles, I would still expect to see a higher default rate on 2 singles as teams wouldn't be able to talk someone into doing that every match if it is a real problem.

What Would a Four Court Format Look Like?

Second, I took a look at how many five court matches end in 3-2 scores, and how many of those would have been 2-2 if court 2 singles had not been played.

For the 2019 championship year, there have been just over 31K 40 & Over team matches that played 2 courts of singles and 3 courts of doubles.  Of these matches:

• 20% ended with 5-0 scores
• 36% ended with 4-1 scores
• 44% ended with 3-2 scores

If all of these matches were to have not had court 2 singles played:

• 27% would have ended with 4-0 scores
• 47% would have ended with 3-1 scores
• 26% would have ended with 2-2 scores

Yes folks, you read that right, more than one quarter of matches would have ended tied.

So, there are tie-breakers in place for this situation, the sets lost, games lost, and percent of games won metrics.  But looking closer, it is really just the first two that matter as the last one will always offer nothing more as if both teams lost the same number of games, they've won the same number to, so percent of games won is a pointless tie-breaker in this situation.

Note that presently on TennisLink, Game Winning Percentage is shown on all team matches it appears, and unfortunately appears to be calculated incorrectly nearly all the time.  Or rather, it is perhaps showing a team's overall GWP and not for the match which seems wrong and not what the regulation calls for.  This is moot for matches that have an odd number of courts played, but for these 2-2 ties, calculating it wrong could identify a winner that isn't the winner.

But carrying on, how often would the two working tie-breakers do the job of determining a winner?  Of the matches ending in 2-2 scores:

• 61% of the time, sets lost would decide the winner, meaning 39% of the time, the tie-breaker would go to games lost
• 97% of the time, sets or games lost would decide the winner, but this leaves 3% of matches for which the rules do not determine a winner.

A number like 3% may not seem like a lot, but in a flight with four sub-flights of 10 teams each playing each team within the sub-flight once, on average between 5 and 6 of these matches would end in ties, sometimes more, sometimes less, with no rule in place to determine a winner.  This seems like a problem to me ...

It is unclear to me what happens then, although I think I've seen some local leagues that state they will give half a win to each team and handle it manually at the end of the regular season since TennisLink doesn't support that.  What would happen in playoffs?  I'm not sure.

Note: I looked at the same stats for 2017 and 2018 and they were basically the same.

How Could This Be Solved?

Assuming changing to an odd number of court format is not an option, how could this be changed?

One is to establish a tie-breaker that will determine a winner.  In this case, the sets lost and games lost tie-breakers are actually not bad (a single head to head match does not introduce the same issues as a round-robin flight does), they just don't always determine a winner.  Obviously the pointless percent of games won should go away so-as to not confuse and require a calculation that TennisLink can't do, but what then instead?

I'd submit the following options if sets lost and games lost leave the teams tied:

• Winner of a specified court - Arguably this should be court 1 singles to give the singles court a little more meaning, but any court could be identified as long as it is the same and part of the rules known in advance.  One could also say to use court 3 doubles so the players there that may be an afterthought have more to play for.
• Have a designated court play a super tie-breaker to decide the win - This could again be court 1 singles or whatever court is identified, alternatively it could be the last court finished to have them just continue.  The challenge here would be for timed matches or situations where there is not time to fit this in.
• Coin flip - A coin flip is random, seems silly to decide it this way, but at least have the rules say to do a coin flip would be better than saying nothing and having 3% of 2-2 matches have no resolution.
• Change the format to Points Per Position (PPP) rather than team record

Elaborating on the last point,  PPP does away with the concept of a team win and just has teams accumulate varying points for winning the different courts all season and have standings based on those points.  So both teams winning 2 courts has no special meaning other than each team accrues the points associated for those courts.  More points are typically assigned to court 1, then 2, and then 3 to try and discourage stacking.

Some areas have been using PPP for local league play for several years now, perhaps the USTA is secretly trying to encourage more areas to adopt it.

Note that PPP works for flights, but not for brackets that are ultimately used at Nationals and most other playoffs as in the case of brackets, you must decide a winner and so must have some way to determine who wins when a 2-2 tie is not decided by the current tie-breakers.

What do you think?  How would you solve this?  Leave a comment here or on Facebook, or drop me an e-mail.

And if you didn't get a chance to vote in the poll regarding the format change, here it is again.

Do you prefer the new 1 singles / 3 doubles or existing 2 singles / 3 doubles format for 40 & Over

1 singles / 3 doubles

2 singles / 3 doubles

Created with SurveyMaker