# A New Way to Think About Break Points

Discussion in 'Tennis Tips/Instruction' started by The Wreck, Feb 16, 2012.

1. ### The WreckSemi-Pro

Joined:
Jun 30, 2008
Messages:
768
Had some thoughts rattling around in my head while watching tennis the other day, and always been interested in statistics, so I made a quick blog (temporary, really) to house my thoughts for the time being. Thought I'd post it here and see what people thought and what suggestions you could provide. Thanks.

http://tennismetrics.wordpress.com/

2. ### kragsterHall of Fame

Joined:
Jul 8, 2011
Messages:
1,658
OP you have some interesting thoughts here. I think your system is great but would end up being somewhat complicated to implement. It did get me thinking though, a simple adjustment from the current system would go a long way - When counting breakpoints, only count 1 breakpoint per game. So if you have a long game that goes back and forth from deuce 20 times, that should only count as 1 breakpoint.

I think this will give a better sense of how many opportunities a player had. If you see someone lose a match with stats like 1 on 20 conversion, you would think that the player had a lot of chances. If those were all in 1 game though, this would be somewhat misleading as really the player had only 1 game that he could have broken.

What we need is distinct game break pt conversion i.e. only 1 breakpoint counted per game in which you had breakpoint opportunities.The one flaw with this system though is that it would not be a good judge of situations when a player choked i.e. those cases when a player has multiple breakpoints in the same game but doesn't even convert one.

Maybe they should have a breakpoint conversion along with a breakgame conversion. That would give us the complete picture.

3. ### Raz11Semi-Pro

Joined:
Sep 17, 2011
Messages:
720
Agreed. Simple and effective if we use both stats.

OP, you do have an interesting stat which is effective but can be a bit complicated. Especially for the casual fans, they would have a hard time understanding this.

4. ### The WreckSemi-Pro

Joined:
Jun 30, 2008
Messages:
768
Understood. It's not practical the way tennis stats are tracked nowadays. As far as the casual fan understanding it, I would hope that it'd be like OPS in baseball. Most people have no idea what OPS is or how its calculated, but they know a higher one usually means a better hitter. Same here. One of my goals (which I havent been able to test yet), is to see how this EBP related to wins (does the higher EBP win 99% of the time, etc.)

5. ### Raz11Semi-Pro

Joined:
Sep 17, 2011
Messages:
720
Theoretically it should work but how would it incorporate tie breaks. Cause some matches are determined by tiebreaks and not breaking of service games.

Example:
Set1:
Player 1 breaks Player 2 but gets broken back. Later breaks again and wins the set.
Set2:
Player 2 breaks Player 1 but gets broken back. Later breaks again and wins the set.
Set3:
Player 1 beats Player 2 in the tie break.

Player 1 had 3 breaks and 1 break back. Same for player 2 but player 1 had more chances.

The EBP if my calculations are correct then player 2 will have the higher EBP but lost the match in the end.

Last edited: Feb 16, 2012
6. ### SentinelBionic Poster

Joined:
Jan 21, 2008
Messages:
48,648
Location:
RR Lyrae
I've always wondered this abt the BP conversion stats (If you had a triple BP and converted on the first it would be 100% right, but then when they show 5 of 8 converted then ...)

How about counting BP's missed rather than converted ?

In the above case you have 0/3 BP's missed. I know this seems to be confusing too, doesn't tell you how many times you did convert when you add them up and have say 5/15 missed. So how many times did you break ?

7. ### sbengteG.O.A.T.

Joined:
Aug 20, 2010
Messages:
12,703
I agree that it is more meaningful to look at these two stats together than just BP conversion.

Have there been instances when a player has won more points on return than the opponent and still broke the opponent fewer times (and lost the match) ?
Does it make sense to look at a stat like
(total points won on return Vs break point conversions) ?

I guess in some way, that stat would be a measure of how well the player played the big points as compared to overall points. Similarly you can look at points won on serve in relation to break points saved. (Again a measure of how well you served on the big points).

It would be interesting to look at this stat in some of the Fedal matches (from before 2008 ) where Fed typically would create a lot of break opportunities but not convert them.

8. ### rosewall4everSemi-Pro

Joined:
Dec 24, 2007
Messages:
501
breakgame will only be useful if it it indicated when the break happened and who in the set served first. a breakgame is evident from set score...

9. ### Fate ArcherHall of Fame

Joined:
May 14, 2009
Messages:
4,108
If I'm understanding this right, the breakgame could be useful, as discussed already, to accurately show how many opportunties a player had in the match, without being overly inflated by the times a player didn't convert a breakpoint in a single game.

We could have a situation where the score is leveled at 4-4, but while player A held easily in all his games, player B faced multiple breakpoints in all his service games.

The breakgame percentage would be 0% (0/4) still, like in the breakpoint conversion (which could be as high on breakpoints missed as say, 0/12), but it would more accurately translate how much a player capitalised or not on his opportunities through the course of a match.

To put it simply, it would indicate how many games you won when you had breakpoints. This may not be implied in the score (two situations where it may not: 1- as in the above instance, a break does not happen. 2 - the other player breaks back to even up the score. In this last case the OP's metric which accounts breaks back may show a bit more insight).

I agree that both stats (breakpoint and breakgame) are complementary to each other, and the use of the breakgame scout could prevent situations where the breakpoint conversion alone would be totally misleading, such as a match where the breakpoint conversion was 9% (1/11 chances), but all those breakpoints came in a single game and thus there was only a single break in this match. It could well happen.

In addition, it would be so much easier interpreting a match through the statistics using these stats coupled, let's take this last example, we can gather that given that the match was decided by a single break, it was all pretty even until this very important game was played, where it took 11 chances for the break to happen and the match to be decided.

TL;DR: Broadcasters should definitely include the breakgame stat on their statistics when showing breakpoint conversions. It can be extremely misleading showing one without the other.

10. ### sureshsBionic Poster

Joined:
Oct 1, 2005
Messages:
45,099
What happened to Breakpoint?

11. ### double barrelsRookie

Joined:
Nov 29, 2011
Messages:
389
Location:
the courts, the bars, or somewhere inbetween
schemathematics....love it

12. ### dominikk1985Legend

Joined:
Feb 11, 2011
Messages:
6,860
this is like in baseball. everybody loves clutch hitting but statistics show that clutch is basically ********.

the guy who has more breakpoints usually wins. and in baseball the guy who hits better will usually hit better with RISP.

also the teams who GIDP the most usually score the most runs. of course GIDP is bad but usually it is just an indication of having a runner on first often which means a lot of runs.