The New Magic Numbers

[johnyandell]

Let's get your thoughts on Craig O'Shannessy's article, "The New Magic Numbers"

[weslas]

In general, I agree with his conclusions, especially with respect to the number of errors in a match and the percentage of points one needs to win (over time).
However, the conclusions on coming to the net may be erroneous.
These days it seems that players come to the net only when the situation seems to be clearly advantageous (e.g. the sneak). As such, most players should have the edge in net points. On the other hand, if one serves-and-volleys almost exclusively (or just likes to come in rather than sit on the baseline), the net point percentage will fall as one would be coming in in less favorable situations.
Otherwise, since winning 55% of points would equal near-GOAT status over a career, one would have to conclude that coming to the net (where >60% of points are won) would lead to automatic greatness.

[klacr]

Great to have this in writing. I saw Craig give a presentation on this earlier in the year during a coaches conference in Washington DC. It was intriguing to see these numbers and more importantly, to see how they are interpreted and implemented.

So critical to understand that just a few points or a few percentages here and there can really create a huge variance in careers and results. Federer, Djokovic and Nadal have won 55% of the total overall points in their career. Andy Murray is at 53%. Take a player like Tomas Berdych, he has been solidly in the top 10 the last 6 years, highest rank of #4 on may 18th 2015, one of the most consistent players on tour (results wise), 99.9% of tour players would happily take his career, almost $25 million in career prize money, yet he's only won 50% of his career points. Thats pretty humbling.

As for the serve and volley statistics, I know Craig was just giving us a taste of what's to come and hopefully will answer your questions, as I know he has the answers having seen his presentation.

Kyle LaCroix USPTA
Boca Raton

[faultsnaces]

I have studied the tennis scoring system extensively (it's an amazing bit of statistical engineering - but they must NEVER switch to no-ad scoring!), and while I agree somewhat with the article, there are far more significant insights to be gained just by looking at the scoring system.

The point margin needed to win is indeed small, but that's a feature of the scoring system, not of the players. Rest assured, if this game was played with basketball-like scoring (fixed time limit, alternate serving), we would see very lopsided outcomes such as 100+ to sub-50. (Challenge to teaching pros: try the experiment with your students; match them up and have them alternate serve for a half hour....)

The tennis scoring system is designed to be able to determine the winner between two players who are very, very closely matched, and also be usable between players who are at completely different levels - in both of these matchups, the match time remains 'reasonable' and gives a very high-confidence answer for which player is 'better'.

Note that the point margin in any tennis match, no matter the skill levels involved, is always small: 24-0 is a perfect - golden - set, and has a 24 point winning margin, but a 6-0 match can be won with a mere 12 point margin, and a set won in a single service break may have a 4 point margin, while a tie break set may have a mere 2 point margin. With players who are reasonably close in playing ability, sets with a 2 or 4 point margin are far, far more common than blowouts of 12+ point margins. As the point margin required to win is generally quite small compared with the total number of points played (golden sets are frightfully rare), the percentage of points won rises quite slowly compared to number of sets & matches won, even for players who win most of the sets & matches they play.

What I see in Craig's analysis is what I always suspected, that competitive play at all levels is actually very, very competitive: players at any given level are very, very close in overall ability - but the jumps between the levels are comparatively huge! This is, to me, a valuable insight and analysis.

The key to understanding tennis scoring is not that it only takes a few points to win, but rather that it takes sequences of points - points won in succession - to win: to win, you must win more point runs than your opponent.

If your opponent wins a point, you NEED to win a point. If you win two in a row, your opponent MUST win two in a row. If you win three in a row, your opponent must match that. If you win two in a row and your opponent then wins three in a row, you MUST win two in a row AGAIN in order to win the game. In short and over simplified, isolated points don't matter; what matters is winning two in a row, three in a row, and four in a row - and similarly (but slightly differently) for games.

And stay away from no-ad; that's the devil's own scoring system. (Tie breaks good, very good; no-ad bad, very very bad.)
-frank

[don_budge]

The Scoring System...Bill Tilden "Maintaining Pressure on Your Opponent"

That was an interesting post faultsnaces. I can tell you have given a lot of thought to the scoring system. The best understanding and bit of writing that I have ever come across regarding the scoring in tennis is from Bill Tilden's "How to Play Better Tennis: A complete guide to technique and tactics".

The truly interesting aspect of the scoring system is that it is built around the psychological aspect and physiological aspect to the game itself. The psychology of playing a match is often dictated around what the score is. The same can be said for the actual technique aspect of each players game and how it is equipped to handle the psychology and the score. Thus you must always know the score...in tennis as well as in life. It is imperative to know where you stand at any given moment and therefore you understand how important it is to be present and in the moment.

Tilden's synopsis of his understanding of the scoring system is found under his Part Three of his book entitled "Match Play Tactics and Tennis Psychology". Chapter 15 is called "Maintaining Pressure on Your Opponent" and it is here that he devotes a section called "Playing to the Score" to analyze point by point the psychology behind the score. I think it is such an amazingly brilliant analysis by Tilden that it makes just about all other explanations unnecessary. But I tend to reduce things to the LCD...the lowest common denominator.

Here are the final three paragraphs of Tilden's comments...

"Many a player plays a set-point or a match-point to him or against him and doesn't realize that anything crucial is going on, because the score does not say so. I mean by that, the final result of many sets and matches hangs upon points which come up early in the first set. Let's take a hypothetical situation which illustrates very nicely the crucial-point and crucial-game theory I have discussed in this chapter.

Two players of almost equal ability are battling in a match that each is very eager to win. Both men are keyed up and inclined to be jittery. Mr. A leads 4-2 and 40-15 on his own service and feels he has the set in hand. On the next point Mr. A reaches the net and Mr. B hits him an easy drive which if Mr. A played carefully he could put away for a win. Instead, without taking proper care he attempts a drop-volley and misses the shot; 40-30 instead of a won game and a 5-2 lead. Annoyed Mr. A goes back and serves a double fault. Deuce. Mr. B, cheered up by these two reprieves given him by Mr. A makes two fine shots and the score is now 3-4, with Mr. B's serve coming up. He is back again in the match, and he eventually wins it. Actually if Mr. A had played his 4-2, 40-15 volley carefully he would have won the game and probably the set and the match.

In my opinion, the match-point was that missed volley, but neither man knew it at the time. So keep alert for the psychological value of certain points, played early and sense the possibility of the equivalent of set-point or match-point coming up for you when you least expect it. Many a match can be turned by taking advantage of the unexpected reprieve, or the lucky break in your favor. Always be on guard against giving your opponent another chance he hasn't earned. Consolidate your gains whenever you pile-up an advantage and be ready to jump in and grab an opening if it's given to you."

[nickw]

I love numbers generally, so no surprise I absolutely loved this article, and am looking forward to the series!

Regards rally length and 70% of rallies ending in 4 shots or less, this just shows the value of having a preferred plan in your mind for your first 2 shots of the rally, for each of the 8 different situations you could face before a point begins. Execute it well, and the point could be yours, or at least you'll be in a dominant position.

67% of points end in errors, I actually thought it was higher, and have always tried to address the importance of error management to players, simply for this reason.

As for net approaches, I think the main message here is to think about all those times that players have a good opportunity to hit and come in, but don't take it. I think players could definitely tap into that 65% success rate just by being clever and choosing the right times to get to the net. Also, it saves the body with shorter rallies, and keeps the opponent guessing.

Nice extract from Tilden's book don_budge, I think that highlights the importance of staying in the present during a match, and playing each point on its merits. 40-0 has exactly the same importance as 40-30, because they both can lead to the game being won. Player's mind should be 'game point' in both situations, once you start thinking 'I got two more if I mess this one up', you're doomed.

Nice post faultsnaces, definitely agree that no-ad scoring must never be used in singles. I like the point about the importance of winning sequences of points, what better way to edge that % up a notch, but of course it takes a consistent mental effort to achieve.

[don_budge]

67% of points end in errors...

67% of points end in errors, I actually thought it was higher, and have always tried to address the importance of error management to players, simply for this reason.

This is the only stat that you need. It is higher...actually.

[lobndropshot]

How do you gather your data?

Hi Craig,

How do you collect your data? Also, have you looked into doubles? I really like working on projects like this and if you need help collecting data at the NTPR levels let me know.

Best,

Lobndropshot

[craigoshannessy]

Hi from Craig

Hi Craig,

How do you collect your data? Also, have you looked into doubles? I really like working on projects like this and if you need help collecting data at the NTPR levels let me know.

Best,

Lobndropshot

Hi there - I LOVE doubles! I have two doubles products on my website. I am very interested in gathering doubles data. Are you already doing it?

Thanks for taking the time to comment.

Best,

Craig

[craigoshannessy]

Hi from Craig O'Shannessy

G'day!

Great to see the replies and discussion about the story. Really glad to be working with John to bring you this brand new data. Please feel free to ask any questions, as I will be checking in here regularly to stay up to date with you guys.

All the very best,

Craig O'Shannessy

[klacr]

Great to have Craig on here. Knew it was just a matter of time.
Not only does Craig know his numbers, he's a teacher at heart. He's clocked in hours on the court and worked in the trenches like many of us on here. He gets it guys.

Kyle LaCroix USPTA
Boca Raton

[bowt]

Craig I saw your presentation at the grand slam conference at the Aus Open and thought it was amazing.

What exercises/drills would you guys do with students to practice serve and return?

I watched one where players serve with different coloured balls but seemed crap to me.

Do you think returners really pick where they are going to hit the return?

[craigoshannessy]

Hi From Craig O'Shannessy

1. Hi Kyle. Thanks for the kind words. Great to be on here following up about the article. Yes, I have certainly done my fair share of hours on the court - running academies, coaching at clubs and also on tour. A lot of that was experimenting to see what would really work, and would not. I experimented a lot with introducing strategy at every single level of the game, to see how it would stick. Guess what - it sticks!!!! These new numbers are only going to get clearer, cleaner, and a lot more of them in the coming months. Cheers mate!

2. Hi bowt - thanks for your nice comments about the Aussie Open presentation (The First 4 Shots).

Thoughts.

a. A have never heard of different color balls to help returners pick up the ball better. I am not sure about that one.

b. With the 8 serve locations, it's really good to put cones down and practice hitting them. In group situations with say 4/court, put the cones first at the body (Position 2.5) and compete court to court knocking them over. First to knock over 3 wins. Then do Positions 1&5 then 4&8 and last round they get to pick.

c. Serve +1. Serve at a target, the coach feeds to "C", and then server hits a run-around forehand to C+

d. Return +1. Serve a weaker 2nd serve, and then feed the Serve +1 ball to C, and have the returner play big mainly to C+ but also to A.

e. Play a set of only serves and returns (1st & 2nd serves). The returner must hit it past the 10&U line to get the point, or a double fault. If the return is in the service box, the server wins the point. If it lands between the service line and the 10&U line, the returner gets a "Re-DO" only ONCE per point. The server gets the same serve back they just hit. If the returner hits it in that region again ON THE SAME POINT, they automatically lose that point. Then it starts fresh again on the next point.

That should get you going!

Best,

Craig

[bowt]

Hey Craig thanks for replying have tried these in yesterdays and today's squad and they were great

you got anymore!?

[johnyandell]

bowt,

just a thought but you could actually go to craig's site and subsidize his efforts...