Why not serve two 1st serves?

[asintu]

Assume your first serve percentage is 50% (or 1/2). If you hit the second serve the same as your first serve your probability of making a double fault would be 1/2 x 1/2 = 1/4 (or 25%)...or 75% that you will get one of the 2 1st serves in...I kinda like the odds. And this is only assuming you have a 50% 1st serve.

For a 60% 1st serve -> 84% chances of getting one in
For a 70% 1st serve -> 91% changes of getting one in (WOW!!!)

So then why not practice only first serves then?

 

[Rickson]

Assume your first serve percentage is 50% (or 1/2). If you hit the second serve the same as your first serve your probability of making a double fault would be 1/2 x 1/2 = 1/4 (or 25%)...or 75% that you will get one of the 2 1st serves in...I kinda like the odds. And this is only assuming you have a 50% 1st serve.

For a 60% 1st serve -> 84% chances of getting one in
For a 70% 1st serve -> 91% changes of getting one in (WOW!!!)

So then why not practice only first serves then?

So according to your calculations, you should always get the second serve in if you make at least 50% of your first serves. OK, let's be for real now. The reason you use a topspin serve instead of a flat bomb for a second serve is the high margin for error. Your calculations are flawed in that you could be a 50% server by missing your first 10 first serves and getting in your next 10. Your calculations seem to make sense on paper, but in tennis terms, they're very inaccurate. I don't find your advice totally insane because I would and have used 2 first serves against opponents who love crushing second serves.

 

[Guga_x]

Altough it might look a totally stupid question it is not, especially for people who have a defective 2nd serve.

I have some kids at my local club who have a 50% on 2nd and about 60 on 1st. So why should they not go with the higher percentage and most effective bomb?

 

[papa]

Assume your first serve percentage is 50% (or 1/2). If you hit the second serve the same as your first serve your probability of making a double fault would be 1/2 x 1/2 = 1/4 (or 25%)...or 75% that you will get one of the 2 1st serves in...I kinda like the odds. And this is only assuming you have a 50% 1st serve.

For a 60% 1st serve -> 84% chances of getting one in
For a 70% 1st serve -> 91% changes of getting one in (WOW!!!)

So then why not practice only first serves then?

The better idea would be to use two "second" serves until you're confident that you can get it in.

I think assuming people can get 50% or more, of their serves in (two in a row) is rather unlikely. Now, even if you give one point away per game thats 25% so the opponent only has to win one of three points (when you don't double fault) to keep even. If he wins two, your can be broken.

 

[misterg]

Assume your first serve percentage is 50% (or 1/2). If you hit the second serve the same as your first serve your probability of making a double fault would be 1/2 x 1/2 = 1/4 (or 25%)...or 75% that you will get one of the 2 1st serves in...I kinda like the odds. And this is only assuming you have a 50% 1st serve.

Assuming you hit your first serve 50% in... it means that you have 50% to get your serve in and another 50% to miss your (first) serve. So assuming you missed your first serve ... when you are attempting to serve for the second time... you have another 50% to get your serve in and 50% to miss your (second) serve and nothing more.

 

[Nuke]

Uh, if you hit two first serves with 50% in, then you'll average a double fault on 25% of the points you play. If you're OK with that, go for it.

 

[Rickson]

Uh, if you hit two first serves with 50% in, then you'll average a double fault on 25% of the points you play. If you're OK with that, go for it.

Even if this calculation were accurate which it isn't, if you double fault 1 out of 4 times per game, you're giving away a point per game without the opponent doing a single thing. Now my calculation also becomes flawed in that the server is supposed to get 2 out of the 4 first serves in so you'd see a double fault at the halfway mark of the next serve game. The conclusion? These calculations are extremely inaccurate and misleading so develop a good topspin serve and don't go for 2 big serves unless you have an extremely weak 2nd serve that your opponent eats up at will; I'd take a chance then, using 2 big serves.

 

[spot]

To decide- you have to say what percentage of games you win on your first serve vs what percentage of games you win on your second serve.

 

[Rickson]

Another thing to remember is that when you go for your big first serve, you have a cushion to fall back on. When you serve the second flat bomb, having no cushion could cause you to lose confidence and double fault from the jitters of having no room for error. I like the idea of going for 2 flat bombs if I'm up 40-0 or 40-15. I'd never go for it at deuce or ad out. I definitely prefer going for 2 flat bombs over 2 topspin serves any day of the weak.

 

[theace21]

Your only as good as your 2nd serve.

 

[Tennis Ball Hitter]

To decide- you have to say what percentage of games you win on your first serve vs what percentage of games you win on your second serve.

This is the only convincing argument in the thread so far.

I have also had similar thoughts especially about Sampras. I also reckon that the chances of getting a first serve in the second time you try it would be greater than the first time you try it.

 

[Tennis Ball Hitter]

Assuming you hit your first serve 50% in... it means that you have 50% to get your serve in and another 50% to miss your (first) serve. So assuming you missed your first serve ... when you are attempting to serve for the second time... you have another 50% to get your serve in and 50% to miss your (second) serve and nothing more.

You are correct in what you say, but the original poster is calculating the chance of not double faulting, you are calculating the chance that the server will fault a first serve on a second serve opportunity ... not the same thing.

 

[Dimethyl Sulfoxide]

You are correct in what you say, but the original poster is calculating the chance of not double faulting, you are calculating the chance that the server will fault a first serve on a second serve opportunity ... not the same thing.

Actually you're a bit incorrect too.

The original poster is calculating the odds of not double faulting assuming the first serve is not hit yet. Once the first serve is hit, the odds of double faulting is 50% again.

 

[Rickson]

You're right, DS, the percentage is once again 50 and not 75.

 

[Tennis Ball Hitter]

could someone explain to me where I went wrong. My probability skills have always been a bit suspect ...

 

[kevhen]

It depends on your opponent and if they can get back your first serves or not and how well they attack your seconds and how good your first serve compared to how bad your second serve is as well as how well you play out points after the serve.

If you have no second serve, no groundstrokes or volley, and a decent first serve, then it may make sense to go with that and take some double faults, but without much of a baseline game you won't break serve much unless your opponent doublefaults alot as well.

 

[Dimethyl Sulfoxide]

Me again. I thought I might add to this thread once more.

If you guys have ever played matches in tournaments, you'll soon realize most players are scared as hell to hit back your serve with any authority. Most players just want to get the ball in play and will usually block or chip. Lots of players love getting fast flat serves and don't handle slower serves well.

If you watch open level players, many chip every return back. It's very important to get the point started to even have a chance at breaking.

The advantage to getting a big second serve is less important when your opponent is chipping the returns. At that rate, you cannot afford to give free points by double faulting. Obviously if they take a big cut at the ball and consistently pressuring your service games, you might have to change tactics. But I've seen many 4.5 and 5.5 win tournaments by spining in both serves two second serves rather than two first serves.

 

[asintu]

some very good points made in this thread. Here's what I have to add:

the calculation I made is only accurate over very many points (maybe at least 4-5 sets) and it also assumes you're a "robot" mentally, which most of the time is not true because you usually don't hit with the same state of mind a second 1st serve when you're 30-40 as opposed to when you're up 40-0. There's also a lot of other variables: like the style of returns of your opponent, your winning percentages off the 1st and second serve etc.
Thing is sometimes it could be a good thing and sometimes it could work in your detriment. The thing is you have to be smart enough to know when to do it and when not to.

 

[Burt Turkoglu]

...I've used 2 first serves playing doubles using the "I"....was very effective at times....except one time I just wasn't feeling it and double faulted 5 times in one game.....is that some kind of record?....lol.....

 

[Rodzilla]

It really depends on how you're serving that day from all aspects. If I'm feeling really confident in my serve that day, sure I'll go for it even on the second try from now and then, but there is a reason for a second serve....

 

[stc9357]

Sometimes when I'm hitting my serve good I will go with 2 first serves. It wouldn't be a good practice to hit first serves on every point because you will double fault a lot. Yes your probability is suspect.

 

[MTChong]

Uhh, yeah guys... You serve a serve, and there is fifty percent chance of it going in, then it will go in or out, flip of a coin. On the next serve, the same percentage is there; it doesn't really matter because it's a new scenario. When you roll a dice once and roll a one, on the next roll, the chances are still the exact same that you will get a one again.

And if you have an effective second serve, you may as well use it. Otherwise, it's a wasted shot.

 

[SageOfDeath]

Uhh, yeah guys... You serve a serve, and there is fifty percent chance of it going in, then it will go in or out, flip of a coin. On the next serve, the same percentage is there; it doesn't really matter because it's a new scenario. When you roll a dice once and roll a one, on the next roll, the chances are still the exact same that you will get a one again.

And if you have an effective second serve, you may as well use it. Otherwise, it's a wasted shot.

exactly, 50% + 50% on your next shot doesn't equal 75%. Its called probability, and the probably doesn't rise to 75%. Its just 50% twice.

 

[Dimethyl Sulfoxide]

Uhh, yeah guys... You serve a serve, and there is fifty percent chance of it going in, then it will go in or out, flip of a coin. On the next serve, the same percentage is there; it doesn't really matter because it's a new scenario. When you roll a dice once and roll a one, on the next roll, the chances are still the exact same that you will get a one again.

And if you have an effective second serve, you may as well use it. Otherwise, it's a wasted shot.

You haven't taken much statistics have you Chong? You are right that if you hit a first serve and fail, the odds of making a 2nd serve is 50% after knowing that you've failed. But what are the odds of getting at least one of those serves in assuming you haven't hit a serve yet.

Let's use your coin example: Let's say I give you a coin and heads means the serve is in, tails means the serve is out. Now I let you flip the coin twice and as long as you get heads you win. What are the odds you will get tails on BOTH rolls? Well it surely isn't 50% now is it? Here's the allowed sequences:

#1: Heads, Heads -- stop after first head since first serve is in
#2: Heads, Tails -- stop after first flip since first serve is in
#3: Tails, Heads -- you got 2nd serve in after failing first serve
#4: Tails, Tails -- you double faulted

Odds of making a serve in for every point played = 75%

 

[All Court]

Exactly. You hit one serve, the chance of it going in according to the original statistic is 50%. On the next serve, getting it in is still 50%. But it's an entire different problem if you're asking the probability of getting one in with two tries. It's 75% then.

But...a 25% chance of double faulting is pretty bad. Even beginning players are ashamed of double faulting once a game. Pro players are disappointed for double faulting at all. Do you realize how many free points you're giving away with that percentage of a double fault?

 

[SageOfDeath]

You haven't taken much statistics have you Chong? You are right that if you hit a first serve and fail, the odds of making a 2nd serve is 50% after knowing that you've failed. But what are the odds of getting at least one of those serves in assuming you haven't hit a serve yet.

Let's use your coin example: Let's say I give you a coin and heads means the serve is in, tails means the serve is out. Now I let you flip the coin twice and as long as you get heads you win. What are the odds you will get tails on BOTH rolls? Well it surely isn't 50% now is it? Here's the allowed sequences:

#1: Heads, Heads -- stop after first head since first serve is in
#2: Heads, Tails -- stop after first flip since first serve is in
#3: Tails, Heads -- you got 2nd serve in after failing first serve
#4: Tails, Tails -- you double faulted

Odds of making a serve in for every point played = 75%

You don't get the point of his post. Key word "probability". Meaning it won't always happen. Yes its somewhat unlikely to get tails both times on a coin toss but if that happened would you be amazed? What about even 3 times?

The point of this whole thread is why not 2 first serves and I think the point of his post was to cover that fact that his calculation of 75% can't be used because its just probability.Not only that but as the original poster said that's only if you are a mental robot.

And even if you get exactly 75% of your serves, that's a lot of double faults.

 

[devilish_duke]

I personally change up my service routine to whatever's comfortable. Sometimes I'll hit 2 first serves, sometime a second and then a first, etc. I prefer to hone my serves so well that I'll have confidence in them and just go out and mix it up.

 

[takeuchi]

Do you realize how many free points you're giving away with that percentage of a double fault?

i think that about sums it up.


might be ok to do it once in a blue moon to mix things up

 

[Dimethyl Sulfoxide]

You don't get the point of his post. Key word "probability". Meaning it won't always happen. Yes its somewhat unlikely to get tails both times on a coin toss but if that happened would you be amazed? What about even 3 times?

The point of this whole thread is why not 2 first serves and I think the point of his post was to cover that fact that his calculation of 75% can't be used because its just probability.Not only that but as the original poster said that's only if you are a mental robot.

And even if you get exactly 75% of your serves, that's a lot of double faults.

I'm not making an argument about hitting two first serves. If you read my earlier posts I was against it and posted examples why not to do it.

However, there is a statistical relevance to this, and Chong's theory relating to coin flips is not correct if you look at it in the big picture. If you serve 50% on average, you will get 75% points started without double faults in the long run.

 

[spot]

Like I said earlier- this dicussion doesn't make any sense until you decide what percentage of the points you win on your first and second serves. The reason that you do a second serve is because you should be able to get it in and if the return isn't crushed you have a 50% chance of winning the point. As opposed to double faulting where you would have a 0% chance of winning the point. If your second serve is so bad that the other person is just killing it then you might have a much lower than 50% chance of winning the point so doing 2 first serves might make sense- but until you decide the percentages this conversation doesn't do any good.

 

[Rickson]

Exactly. You hit one serve, the chance of it going in according to the original statistic is 50%. On the next serve, getting it in is still 50%. But it's an entire different problem if you're asking the probability of getting one in with two tries. It's 75% then.

But...a 25% chance of double faulting is pretty bad. Even beginning players are ashamed of double faulting once a game. Pro players are disappointed for double faulting at all. Do you realize how many free points you're giving away with that percentage of a double fault?

Are you sure about your statistics, Kana?

 

[MTChong]

I'm not making an argument about hitting two first serves. If you read my earlier posts I was against it and posted examples why not to do it.

However, there is a statistical relevance to this, and Chong's theory relating to coin flips is not correct if you look at it in the big picture. If you serve 50% on average, you will get 75% points started without double faults in the long run.

I'm taking Stats right now and only a junior and eventually we'll get to probability, but I know that it's 75%, but I would count both serves as a whole new set of data because in reality, when you're out there, can you really just pull off a second serve with a first serve as if it was a first serve? You really have to factor in the intangibles, and because that isn't possible, I took the liberty to say it's two entirely different scenarios. And of course, you misread my whole thing about the coin. But I think we're talking about two different things here; you about the total, and mine, individually. As someone stated, we aren't machines.

 

[joe sch]

Goran Ivanisevic !

 

[tennis4losers]

Goran Ivanisevic !

I always thought he had two 2nd serves. :neutral:

 

[x Southpaw x]

In singles it's good to aim for 60% first serve, 90+% second serve chance.
In doubles it's good to aim for 70+% first serve, 85+% second serve chance.

These are ideal numbers for high-level play. Difference between singles and doubles: Singles you want no double faults at all even if you use an attackable serve, since you can work out the point. Doubles however is very offensive, and you want that first serve in, so your partner can finish the point off. Losing the offensive on a serve might mean losing the point.

BUT. If you have numbers like the OP said 50/50... suggest you go for a whole bunch of practice serving sessions...
Medium to High number of double-faults = You rot.

 

[fleabitten]

Actually you're a bit incorrect too.

The original poster is calculating the odds of not double faulting assuming the first serve is not hit yet. Once the first serve is hit, the odds of double faulting is 50% again.

And you're incorrect... Because the first serve may have gone in!
What you should've said is "Once the first serve is missed, the odds of double faulting is 50% again."

 

[fleabitten]

You haven't taken much statistics have you Chong?....

#1: Heads, Heads -- stop after first head since first serve is in
#2: Heads, Tails -- stop after first flip since first serve is in
#3: Tails, Heads -- you got 2nd serve in after failing first serve
#4: Tails, Tails -- you double faulted

Odds of making a serve in for every point played = 75%

Thanks DS! It seems so obvious to me. People are getting hung up on the second serve scenario. But the original post clearly understands the probability is with regards to a two-serve scenario. Math is easy until it gets emotional.

 

[jt60312]

Like I said earlier- this dicussion doesn't make any sense until you decide what percentage of the points you win on your first and second serves. The reason that you do a second serve is because you should be able to get it in and if the return isn't crushed you have a 50% chance of winning the point. As opposed to double faulting where you would have a 0% chance of winning the point. If your second serve is so bad that the other person is just killing it then you might have a much lower than 50% chance of winning the point so doing 2 first serves might make sense- but until you decide the percentages this conversation doesn't do any good.

Not that this thread needed another mathematical error (the original poster was correct as already noted), but what's the deal with the 50% chance of winning a point if you get the second serve. That doesn't make any sense at all. I won't insult your intelligence by explaining why. But I'll give a hint to Spot: imagine you're playing Fededer. Now recalculate your odds of winnig the point if you get your second serve in.

 

[spot]

jt- we are talking about a theoretical second serve where neither player gets an advantage. The ball is in play with no player having any advantage so the odds of both players winning the point is 50%. Once you are talking about either a very strong second serve or a very weak second serve then the odds change which is what I was saying the whole time.

 

[ktncnttl]

I think if your first serve % is truely 50 then if you serve it twice you have a 75% chance to get it in at least once, you are right about that. It is like the chance of getting a head at least once when you flip a coin twice. However, first of all 75% is way too low. Secondly, it is inevitable that your first serve % will drop significantly on the second trial because of added pressure etc. Therefore it is impossible to have a two 50% first serve, more like 50% and 30% so on and so forth. Unless your first serve is as big as ARods I wouldn't recommend it.

 

[jt60312]

jt- we are talking about a theoretical second serve where neither player gets an advantage. The ball is in play with no player having any advantage so the odds of both players winning the point is 50%. Once you are talking about either a very strong second serve or a very weak second serve then the odds change which is what I was saying the whole time.

Spot- I know what you are saying. But the original post (and the point of this whole thread) is how many times you would double fault. It does not even bring up how often you would win the point.

I agree that until you take into account "what percentage of the points you win on your first and second serves" you cannot really not begin to find out how helpful this strategy would be to you. If you could somehow come up with all of the statistics for yourself and for your opponent (including those separate from the serve), you could make some interesting decisions about your game strategy.

Of course, most of us cannot do this.

 

[spot]

To me the only question that matters is whether its a good idea to have a second serve or to just use your first serve twice. You can feel free to ignore that question if you want to

 

[jt60312]

No, I definitely think that it is a good idea to have a second serve. Essentially, I agree with you.

 

[Rickson]

But I'll give a hint to Spot: imagine you're playing Fededer. Now recalculate your odds of winnig the point if you get your second serve in.

Try winning the point against Federer even if you get your first serve in.

 

[FiveO]

Interesting mind game.

However, the reality is that the faults and double faults won't occur symetrically. Look at the 50% scenario again. While satisfying that 25% or 1 in 4, rate of failure, randomness dictates the double faults will not be spaced evenly. It won't be 1-2-3-miss, 1-2-3-miss, etc.. Randomness states that those misses may come together in pairs or bunches. How about at 15-30.

This discussion involves reversed reasoning. Its like getting lost in follow throughs to the point of distraction and losing sight that a proper follow through is a RESULT of the proper balance and mechanics that precede it. You can also perform everything incorrectly and artificially force a proper follow through. But to what end?

Serve percentages are a RESULT and are success/failure indicators to be analyzed in the post mortem of a match. The backward reasoning of using a mathematical formula to take what are, in reality, poor serve percentages and somehow transform them into some kind of successful number is frankly, unsound. Like turning iron into gold or like something a snake oil salesman might peddle. The expression "you're only as good as you're second serve" means that you need a strong and reliable second serve that you make near all the time and is effective enough to keep you at least neutral in the point. One shouldn't depend on the law of probability nor a coin flip to decide outcomes on the court.

If this thread is intended as an intellectual exercise, enjoy. But, anyone flirting with this "idea" as a viable approach to play would be better served signing off, shutting down, grabbing a hopper of balls and getting to the job of adding spin to their serves via racquet head speed.

Make your second serves strong through practice of sound technique and you'll never look back.

First rule of tennis....hold serve.

 

[misterg]

Theoretically (looking the big picture...) it mean's that you can double-fault for the entire match or even for two, three or more matches in a row befor get into the ace-mode...

 

[JSummers]

Let's assume these typical percentages:

First serve percentage: 60%
Second serve percentage: 90%
First serve won percentage: 60%
Second serve won percentage: 40%

Doing 2 first serve gives winning percentage of:
(first serve in: 60%*60%)+(first serve again: (1-60%)*60%*60%)= 50.4%

Doing first, then second serve:
(first serve in: 60%*60%)+(second serve in: (1-60%)*90%*40%)= 50.4%

So they are the same in this scenario and the balance point. Depending on whether the first or second serve winning percentage is higher than assumed, it favors that side.

e.g. 70% first serve winning & 40% second, you should do 2 first serves
e.g. 60% first serve winning & 50% second, you should do first then second.

 

[kevhen]

With 25% double faults, you must win 2/3 of the points you do get your serve in to win 50% of the points overall. Generally when you serve, you want to be winning well over 50% of the points. There are times when you want to go for 2 first serves, like when you are way up or maybe even way down and your opponent is starting to sneak well inside the baseline as a surprise tactic.

 

[tom-selleck]

With 25% double faults, you must win 2/3 of the points you do get your serve in to win 50% of the points overall. Generally when you serve, you want to be winning well over 50% of the points. There are times when you want to go for 2 first serves, like when you are way up or maybe even way down and your opponent is starting to sneak well inside the baseline as a surprise tactic.

very well put!.... double faults are death unless you have a huge serve where you're getting 40%-50% of your points "free" i.e. uncontested or non-contested.

i think you can hit both serves hard with moderate spin, but you'd probably be taking something off your first serve.

 

[misterg]

As said Brad Gilbert in his book Winning ugly, it goes something like this: "every time you attempt to serve an ace - look in the mirror, do you see Pete Sampras?"... If the answer is yes, got for it, otherwise put the serve safely in.