The Book

[MARCPELLETIER]

Just got my edition of "The Book: playing the percentages in baseball".

http://www.amazon.com/gp/product/B000EW872M/qid=1142267634/sr=1-1/ref=sr_1_1/104-2078050-9365556?s=books&v=glance&n=283155

This is a great tool for every diehard Strato fan---and for baseball fans as well. Over a hundred tables showing real-life and theoretical calculations about how to win the game.

I'll put in here some of the thoughts found in the book, but there is so much tables that I think it's worth having the book.

Perhaps the most controversial claim so far: in low-run scoring environment---say facing a good pitching in the national league---bunting early in the game appears after all as a good strategy.

(the bones of the argument leading to this conclusion is that 1- on average, 11% of all bunt attempts result in singles, and that 2- sometimes, after a bunt attempt that results in a foul ball, you can still let the hitter loose. On face value, these arguments seem to apply to Strat as well: very good bunters (bunt A) have 11% of chances of having a single out of a bunt attempt, and in Strat, like in real baseball, some bunt attempts result in the famous 1&2 and retry, where you can still decide to go with a bunt or let the hitter loose).

[MARCPELLETIER]

BTW,

they too conclude in their book that the #3 hitter of your line-up should be your [b:d33ff9c5b2][i:d33ff9c5b2]fifth[/i:d33ff9c5b2][/b:d33ff9c5b2] best hitter, after #4, #1, #2, #5.

This is now the third straight analyst reflecting on this subject that concludes this way.

[cummings2]

So the Best hitter bats cleanup, the second best leadoff, the third best in the 2 hole and the fourth best bats fifth? or the other way around?

I remember when Adam Kennedy first got to the Angels in the Mo Vaughn trade that Scioscia batted him 2nd for a good chunk of the season, since then he's been batting 9th. Then last year if memory serves me well Torre had Cano in the 2 spot as well. Wonder what the rationale was there. Also, last season when Cabrera (Orlando) was struggling heavily Scioscia moved him from the bottom of the order to 2nd and he found his swing. I wonder if the 2 hole is for whatever reason the most beneficial to the hitter, considering that the leadoff would be OB and the guys on Deck and in the Hole would be good enough to make the pitcher have to pitch to the 2nd bat but in a more "pressured" way.

Thanks for sharing Lucky, will have to go to Amazon...through Moose's site of course :wink:

Is there any mention in regards to the nature of the outs the hitters give so as to determine if they're the "best" i.e. flyballs, groundouts, DPs... or is it the simple AVG/OPS way.

Cool stuff.

[MARCPELLETIER]

Actually, here is their final recommandation:

[quote:61deb8943f]The Book says:

Your three best hitters should bat somewhere in the #1 #2 and #4 slots. Your fourth and fifth should occupy the #3 and #5 slots. The #1 and #2 slots will have players with more walks than in the #4 and #5 slots. [/quote:61deb8943f]

There is like 100 "the Book says" comment just like that throughout the book.

[JAYDINGESS1]

Lucky,

Best is a relative term to what specific number or attribute you are looking for. IE, would you say that Pujols, Eck, Walker (when he was healthy), Edmonds or Rolen were the best? Depends on what defines best. However, take whatever category you want to look at and Albert is the best. This may be a slanted opinion but you want OBP, HR, RBI, Runs, BB/K percentage? I need more info before I would buy in to your and the book's definition.

[Mean Dean]

Bunting is definitely a better strategy in real baseball than in SOM. Two big factors: 1) In real life, there can be an error on a bunt; this never happens in SOM. 2) In real life, if you never bunt, the infield can play further back and expand the lateral range of balls they can get to. So you need to bunt sometimes, or at least be a credible threat to do it, in order to keep the defense honest. SOM doesn't replicate this.

That said, it's still possible that the SOM bunt is underused. But I dunno if you can base conclusions about that off real-life research, because the SOM bunt < the real bunt, unfortunately.

Re: the #3 hitter: If you think about it, it's very logical. The most important hitter in any given inning is the leadoff hitter. And #3 leading off an inning is relatively rare. I dunno about literally batting your fifth-best overall hitter there -- I think the guy should at least have power, since you're giving him more plate appearances, he has to be contributing something -- but, maybe your fifth-best [i:5f95d185f1]OBP[/i:5f95d185f1] guy, if he is indeed a slugger. I totally buy that.

Honestly, if I only had a handful of good hitters, I wouldn't get so fancy; I think it stops making sense at that point. In that case, I would in all probability want one of the good hitters to hit 3rd. That way, he gets more plate appearances, plus his production interacts more with the other good hitters, which produces more runs than spreading them out.

But in the modern game -- where there aren't many glove men anymore, and most teams are fielding eight or nine guys who at least hit a little bit -- a low OBP/high SLG hitter 3rd makes a lot of sense to me, if you do have a player like that.

[MARCPELLETIER]

Dean,

that's a very good point. In fact, I disagree with the logic that you have to put your [i:38cbd4977c]fifth[/i:38cbd4977c] best hitter. In an exchange of emails with the authors of the book, I mentioned that they did not carefully interpreted their own data. What their data are really saying, I said, is not that you should put your best hitter in the first or fourth slot and your fifth best hitter in the third slot, but rather that you should put the player with the best walk ratio at the first slot and your player with the most extreme low OBP-high HR profile (as long as they are good hitters).

The reason for this is because walks have a relative weight that is much more important for the first slot than for any other slots, and because homeruns have a relative weight (compared to other events) that have more weights for the third hitter than for any other slots.

To back-up my claim, I provided an example. In that example, I had Ken Griffey hitting 50 hrs in 500 PA, with a low on-base, but still with the best OPS of the team. In the example that I built, the line-up was maximized with Griffey hitting third--not fourth--despite the fact that he had the best OPS.

My own conclusion is that the team, overall, gets more production is you concentrate if you can put a hitter who will hit a lot of hrs despite low on-base in third slot and a hitter with a lot of walks and on-base in the first slot.

The authors acknowledged :

[quote:38cbd4977c]
I think my general rules stand, but I agree that you'd
have to plug in the numbers to get a more accurate
account. ...

Really what you have to do is figure out each player's
LWTS at each batting order, relative to his own talent
levels. Then you simply maximize each of these
numbers.[/quote:38cbd4977c]

[ANDYCOCHRANE]

This information about the book is similar to the information I posted last week re the Lineup Optimisation tool on Baseball Musings. I also saw a thread about The Book there and having #3 hitter be your 5th best OPS guy. I haven't had much time to see if it works in Strato. I was going to work out my L/R OPS from the cards and put together a lineup close to The Book's sayings.

Problem is, what to do about clutch and DP numbers when settling a lineup? I.e. do you want your 4th best OPS man hitting #5 in Strato if his clutch figures are poor, similar with # 2 hitter with lots of DPs

[ANDYCOCHRANE]

I have recently worked out a lineup vs RHP for one of my Strato teams by what "The Book" says,and this put Helton in the #4 spot - (Best OPS.) However, when I put my team through the Baseball Musings Lineup Analysis tool it gave the best position for Helton as #2 every time and H.Matsui as #4. This has led to a bit of confusion as I had always put Helton #3 previously. The lineup analysis tool also always put Brady Clark in leadoff whilst "The Book" would have him at #5, (4th best OPs vs RHP) in my team and H.Matsui leading of, (2nd best OPS vs RHP)

Seems confusion between the experts as where to put the best hitters. The one constant is that they both say that the best hitter should NOT be put in #3.

[MARCPELLETIER]

[quote:7d0cb0fabf]Problem is, what to do about clutch and DP numbers when settling a lineup[/quote:7d0cb0fabf]

It is obvious that you have to factor them in. The thing is, clutch and DP in Strat and in real life might not be equivalent.

After looking at millions of data, the "Book" says that there appears to exist a small clutch effect, but the effect is rather quite small, and probably are not important to take care.

In strat, clutch is real and important. There is no doubt for example that, for similar OPS, you should put your best clutch 5th and your worst 3rd. But of course, that's an easier problem than solving what to do when your best OPS is also your worst clutch.

Speed is another variable that plays differently in Strat and in real-life.

All this being said, the tables of the Book remain highly valuable, especially if you coach yourself your Strat team (in a face-to-face league, for example, or on internet play), because, with a little computation, you can easily know who to coach in countless situations.


To give one example, it is commonly known that you shouldn't try to steal unless you have 70-75% of success (so when the lead stealing number is 14 or 15). When you really need a steal, sometimes owners will go down to 60%-65%.

But these numbers are based on averages. In fact, on some occasions, it's not worth trying to steal unless you have 85% of success, while in others it is worth trying even at 55% (such that the overall average is something around 70%-75%). The "Book" provides you with the data to know exactly what situations only require 55%, and what situations requires 90%. Thus, even though you end up with an average of 70% success rate---just like the other owners who have learnt sabermetrics--you run creation values could be substantially higher.