Games » Cleveland Indians
Aug26Once again, walks lose a game
Lee Judge
None
I’m no genius (just ask my wife), but the first rule of winning baseball is 1.) THROW STRIKES! (The second rule is to make the routine play and the third rule is to cheat if you have to, but let’s concentrate on rule #1 for now.) Geez, there are only three bases and if you walk three guys in an inning that means you can’t make any other mistakes.
Unfortunately, Felipe Paulino also gave up two hits in the 7th (Hey! The hardest inning in baseball!), so the Royals were two runners over the limit. You can live with two hits, but not if you mix in three walks. As usual, I’m a little hazy on actual facts, but I’m pretty sure the Royals either lead the league, all of baseball or the Milky Way Galaxy in walks.
When you’re looking for fixes, this is a pretty simple one: throw strikes.
Some specifics
Tim Collins has got great stuff, but walks too many people. No matter how great your stuff is, if you can’t get that stuff in the zone, it’s not doing you much good.
Alex Gordon had his 20th outfield assist and here’s how he’s doing it: the Royals play a shallow outfield so Alex is already closer to home plate than normal. But his experience as an infielder is also playing a role. Gordo is not afraid to charge the ball hard and that cuts down the distance he has to throw. He’s also got a quick release and an accurate arm and, once again, all those years at third come in to play. Outfield assists make everyone assume the outfielder has a great arm, but positioning, charging the ball and a quick release make the arm even better.
The Royals have now thrown 22 runners out at the plate. For every great throw, remember, there has to be a great catch and tag to complete the play. Salvador Perez was on the receiving end of Gordo’s throw and got the out, but once again (just like he did against the Red Sox) Perez rolled the left shoulder down and towards the runner to protect himself. Don’t be surprised if the Royals do not get the call from an umpire that believes Salvador put his body between the runner and the mitt and never made the tag.
Jason Kendall told me throws from lefties tend to tail towards the arm side. Melky Cabrera’s throw to home did just that and that’s why Carlos Santana scored on that play at home in the 7th.
I’ve come to believe there is only one kind of true leadership: leadership by example. Everything else is someone trying to talk other people into doing something they don’t want to do themselves. So Jeff Francoeur and Alex Gordon being in the lineup is leadership. Both are banged up, but came to play. That sends a message to the rest of the team: you better really be hurt before asking out of a game.
I’m reading the entire Patrick O’Brian series of books this summer (the guy who wrote “Master and Commander”). Somewhere in all those books is the line, “the judgment of a ship’s crew or an English village is never wrong.” (And if it’s not in there, it should be.) The point is that people who live so closely together see each other at their best and worst, there are few secrets. When I read that line, I thought they should add a baseball team to the list: players live together for months and get to know each other better than any outsider ever could. And that brings us to Alex Gordon: I’ve asked people who the clubhouse leaders are and everyone I talked to mentioned Alex Gordon without me bringing up his name. They all say he’s quiet, but leads by example. Last night proved it.
Any time you see a batter hit into a shift (and we’ve seen several instances in the last four games) you’re seeing good baseball. Some coach studied spray charts, positioned the infield correctly and a pitcher executed the proper pitch. That routine 4-3 is anything but.
The other philosophy
Talented players make smart managers. I was lucky enough to have very talented players when I was managing and nobody was more talented than Danny Jackson. Danny had pitched in three World Series (Royals, Reds and Phillies) and had forgotten more about pitching than I’ll ever know. When Danny first got to the team and I asked him how he wanted the defense set up behind him, he said, “Put ‘em in the bare spots, they’re there for a reason.”
After I got done laughing I asked him what he meant.
Danny said he didn’t know what he wanted to throw next until he saw the hitter react to his last pitch. One pitch followed another in a logical sequence. A hitter that turned on an inside pitch and pulled it foul, was now set up for a pitch away and so on. The last thing Danny wanted was to have a hitter set up for a pitch and not be able to throw that pitch because the defense was out of position.
So he wanted the defense straight up. The middle infielders can see the signs and pass them along to the corner infielders and everybody can see the catcher move and shade that way as the pitch is delivered, but both philosophies count on the pitcher being able to hit his spots.
Either way, we’re seeing a plan in action.
Aggressive base running
I received an email from a reader who wondered whether the Royals’ aggressive base running was paying off. He thought Jeff Francoeur’s attempted steal of third in this game was misguided and wasted one of the Royals’ 27 outs. I’m a big fan of the steal and aggressive base running, so I told him I’d respond here on the site, because it’s an interesting debate. So here goes:
The Royals are on a pace to score more runs than last year, so by that measurement, the aggressive base running is paying off. The won-loss record isn’t much better, but most observers point to the pitching as the main culprit.
Aggressive base running also pays off in ways that are hard to measure. Hitters get more fastballs, pitchers have to use the slide step and are distracted, outfielders rush throws and defenses have to play out of position to make sure they can cover a base the runner tries to steal. Aggressive base running pays off in numerous ways that aren’t apparent when you look at a box score.
And, finally, you can’t really save one of those precious outs (although, Earl Weaver agreed with the reader, so you may want to take this part of the argument with a grain of salt.) Doing nothing may appear to conserve outs, but it really doesn’t .
Something is going to happen next. You can’t just sit there and freeze time. A manager has to place a bet on what’s going to happen next before he knows the outcome. So if you’re trying to get Francoeur to third (and with one out it was not a sure bet Frenchy would be sent on a base hit) do you place your money on Giavotella’s .263 batting average or Francoeur’s 70 percent success rate while stealing bases?
Twenty-six percent vs. 70 percent: is that really a hard decision? I know it looked bad because Johnny got a hit that might have scored Jeff, but the Royals have to place that bet before they know the outcome. Hitters mainly make outs. A great hitter makes outs 70 percent of the time. Sitting on your hands isn’t preserving precious runs, it only appears that way if you don’t think it through.
Refusing to take a better bet because someone will be critical of you if it doesn’t work is not good baseball. It’s covering your rear end.
Hosmer
Paulino
Perez
Not sure if it means anything, but Felipe Paulino is at 111 innings now, most in his career. He tends to throw a lot of pitches, 120 tonight in 6+ innings, which may be a reason he got gassed late and gave up the two walks and two singles. Seems a consistent thread this year, high pitch counts and relatively low innings. If the staff could average one more quality inning per game I think it would make a big difference in results.
I like your use of percentages in the Frenchy steal attempt: 26% v 70%.
I'm not sure I get what you mean by 26% vs 70%. Even if Frenchy is successfull you still need something else to score him (hit, wild pitch, sac fly). Also, if he doesn't steal and there was only one out, you have a chance to score him with a Giavotella or Perez, and if one of them walks, even with the next hitter. By your logic, they should steal every time because even the worst stolen base percentages are better than the best batting averages. Heck, even Billy Butler has a 50% success rate in his career!
Jim: I know they consider Paulino a horse and they hope he goes deep in every game. When he pitches it's a night Ned might not have to get into the pen early.
But that's the gamble: if you try to get Felipe through seven he's getting gassed towards the end because of all the pitches and you better keep an eye on him.
Last night, with only a one-run lead, there was just no margin for error.
Every pitcher would do themselves a favor by throwing strikes early in the count and forcing the action before they've thrown six pitches. If you're struggling to hit corners and you're going to have to lay a fastball in there, why not do it before the hitter knows you're going to?
The pitch counts and innings are starting to pile up for all these guys.
Great work. This is still the first bookmark I click every morning.
Roger: OK, let's run through that inning again: Frenchy leads off with a double. It's rare to steal third with nobody out, you've got all three outs available to get the run in. (It's done, but you better be a 100% lock to make it and few things are 100%.)
The first guy to come to the plate after Jeff's double was Moustakas. He's got two sac bunts so, even though that's not his game, you could ask him to lay one down. But Moose has been hot so they let him swing away. His job was to pull the ball and make sure Jeff was on third or in when his at-bat was over. Moose strikes out, Jeff's still at second,so it's on to plan B.
The score was 0-0 so one run is meaningful, but it's early, so Cleveland is unlikely to bring their infield in. Get Frenchy to third and Giavotella's odds of driving him in soar. Johnny's batting average is .263, but he puts the ball in play 85% of the time.
Those are the odds the Royals bet on: Frenchy steals successfully 70% of the time and Johnny gets the ball in play 85% of the time.
Leave Jeff at second and your betting on batting averages of .263 (Giavotella), .245 (Perez) and .252 (Getz). I like the 70%/85% bet better.
Admittedly, these are averages. Their match-up odds in this particular game against this particular team would be different. And that explains why you don't steal Billy.
He's 1-2 in stolen bases and I can't remember if it was the back end of a double steal or a hit and run where someone dropped the ball or had a heart attack when they saw him take off, but his actual odds of making it on a straight steal or just about zero. Which is why averages are only a guideline for managing decisions.
And I do think the steal should be used more often: it's one of the best bets available. The Royals want to be successful 75%, Doug Sisson wants to be successful 100% of the time.
They've got everything timed out and when a runner takes off the math says they can make it. Reality (a bad jump, a pitcher who hardly ever slide steps does, etc.) intrudes about 30% of the time.
While reality prevents you from stealing everybody in every situation, not stealing with some people may look better, but it's actually betting on worse odds.
Roger: OK, let's run through that inning again: Frenchy leads off with a double. It's rare to steal third with nobody out, you've got all three outs available to get the run in. (It's done, but you better be a 100% lock to make it and few things are 100%.)
The first guy to come to the plate after Jeff's double was Moustakas. He's got two sac bunts so, even though that's not his game, you could ask him to lay one down. But Moose has been hot so they let him swing away. His job was to pull the ball and make sure Jeff was on third or in when his at-bat was over. Moose strikes out, Jeff's still at second,so it's on to plan B.
The score was 0-0 so one run is meaningful, but it's early, so Cleveland is unlikely to bring their infield in. Get Frenchy to third and Giavotella's odds of driving him in soar. Johnny's batting average is .263, but he puts the ball in play 85% of the time.
Those are the odds the Royals bet on: Frenchy steals successfully 70% of the time and Johnny gets the ball in play 85% of the time.
Leave Jeff at second and your betting on batting averages of .263 (Giavotella), .245 (Perez) and .252 (Getz). I like the 70%/85% bet better.
Admittedly, these are averages. Their match-up odds in this particular game against this particular team would be different. And that explains why you don't steal Billy.
He's 1-2 in stolen bases and I can't remember if it was the back end of a double steal or a hit and run where someone dropped the ball or had a heart attack when they saw him take off, but his actual odds of making it on a straight steal or just about zero. Which is why averages are only a guideline for managing decisions.
And I do think the steal should be used more often: it's one of the best bets available. The Royals want to be successful 75%, Doug Sisson wants to be successful 100% of the time.
They've got everything timed out and when a runner takes off the math says they can make it. Reality (a bad jump, a pitcher who hardly ever slide steps does, etc.) intrudes about 30% of the time.
While reality prevents you from stealing everybody in every situation, not stealing with some people may look better, but it's actually betting on worse odds.
Man, I liked that post so much I put it on here twice, sorry.
And, Tim, thanks very much. It's a lot of work, but hearing from people who enjoy it, helps make it worthwhile.
What I think you are not taking into account is the fact that when the steal fails, you have guaranteed that he won't score. On the 70% of the time he is successful you have only given him a better chance to score. There are several results that would have scored him from 2nd or 3rd (extra base hit, some singles, back to back hits), while there are several results that only score him from 3rd. I not saying the the steal attempt was necessarily a bad move, but it's not a no-brainer as you imply. In the early season when we were making it 80% of the time, the running game was definitely helping. Lately we have been averaging in the upper 60's. I'm not saying to stop stealing as a team, and I'm not saying to stop any particular player from every stealing. It just may be wise to tinker a little with how often some of the guys run to see if we can get our success rate higher.
Lee, I think you are right on the "steal" issue, but even more right on the issue of walks. The Royal pitchers are driving me nuts (okay, it's just a short putt and not a drive)putting men on for free, especially in tight situations. Has anyone kept track of the percentage of balls/strikes by Royal pitchers and the League in general? Ned needs to make this a priority for pitchers next year the way he emphasized the running game for field players this year.
Roger: Yes, the steal failing means 30% of the time you will not score that run, but concentrating solely on that number ignores the fact that the odds of scoring that run were even worse if you don't attempt the steal.
That situation is not always a 'no-brainer', but with Frenchy on second and the guys coming to the plate, it was close to being a no-brainer. What I don't know are the match-up numbers that may have changed the odds.
And the Royals do nothing but tinker with the running game. Doug Sisson spends hours studying video of pitchers to give the runners the best chance. Sending the runner is not an overall policy done without regard to situation, having Frenchy steal meant they had the odds in their favor.
It's broken down by pitcher, catcher, runner and count. Pitchers are slower to the plate in some counts than others.
The Royals wouldn't have sent the Francoeur if they did not believe the odds favored that move over standing pat.
Joel: I'm sure someone has the percentage of strikes the Royals pitchers throw, they announce the numbers after every game. That number would probably be available on one of the metric sites.
But the number can't be good, the Royals are way up there in walks. It's really the most basic thing in baseball, you've got to throw strikes.
And the earlier you do it in the count, the better. You can get away with a fastball 0-0 that will get crushed 2-0, 2-0 or 3-1 because the batter is expecting it. It's pretty basic baseball.
A couple of quick thoughts ... the Royals aggressive baserunning has been, according to Fangraphs' analysis, almost exactly neutral this season. They do a pretty comprehensive study of baserunning, including steals and caught stealing, pick-offs, going first to third on singles, first to home on doubles, and so forth. And the Royals baserunning has netted us 0.1 additional runs on the season so far compared to average. Now last season we cost ourselves several runs, so getting back to zero is a step in the right direction. But right now, the extra outs are killing the value from the extra bases.
I think Yost needs to be called out for bringing in Collins. Collins can be effective, but the simple fact is that he has the worst walk percentage on the team by a wide, wide margin. That is another managerial decision that cost a game. This season, it doesn't really matter. But when we become competitive down the road, this kind of loss will really sting.
Oh, and I love the Patrick O'Brian novels. The first three in particular are outstanding, and they are all enjoyable. Maturin is one of my favorite characters in literature. Good times.
"I'm not sure I get what you mean by 26% vs 70%. Even if Frenchy is successfull you still need something else to score him (hit, wild pitch, sac fly)."
Roger, I'm going to butt in on this because I've dealt with a few statheads who are in anguish over Frenchy sometimes getting thrown out. The idea is that a runner on 3rd with one out can score on a passed ball, a wild pitch, a balk, a sacrifice bunt, a sacrifice fly, a grounder to the right side, an error, or a hit. As Gio strikes out 15% of the time, that leaves an 85% chance that he either draws a walk or puts the ball in play. Gio is a bat control guy, so capable of the ground ball the other way. If Frenchy is at 2nd you have a .26% chance of Gio getting a hit with about a 75% chance of Frenchy scoring on it, about 4% of drawing a walk, then if Gio doesn't score the runner, Perez comes up at .245 or Moose at .220.
The hitters following Frenchy are much of the reason for his aggressiveness on the bases. If he was followed in the line up by Melky, Billy, Hoz, and Frenchy he wouldn't need to steal as much or try to take the extra base, but facing Ubaldo, whom the Royals weren't doing much with anyway, putting the runner on third with one out was worth the risk.
Curtis, have had a few discussions on fBsr and I question the usefulness of a stat that shows Billy a better baserunner than Frenchy. Bsr is overly weighted on caught stealing or advancing and is, of course, totally unable to quantify context of a coach's decision to send a runner. Billy doesn't get thrown out because he doesn't get sent, but that also means that if he singles it will usually take three more created bases after him to score a run while a hitter who turns a single into a hustle double or steal can then score on one more single.
Shouldn't the percentages take into acount Frency' success at stealing THIRD base, not total steal success? Third, unless things have changed, it harder to steal than 2nd b/c the throw is shorter.
Just asking.... ;)
Lee: I must not be making myself clear. I have not said that it was a bad decision to steal. My only point was that I don't think it's automatic that you steal in that situation. I think if you ask Ned Yost he would tell you that they would not always steal in that situation. Your defense of the play was that it was an easy call because you have a 70% success rate vs 26% chance of the next batter getting a hit. As everyone in this discussion has mentioned, it is much more complicated than that.
They jinxed Paulino. They talked about that opponents BA with two out. And if I recall, they got a few hits with two outs last night. Also, 0-2 on Hannahan, ugh. Still, leaving the bases loaded earlier hurt big time.
"Third, unless things have changed, it harder to steal than 2nd b/c the throw is shorter."
I've always heard that 3rd is easier because the runner gets a better jump and with an RH hitter it is easier yet, the catcher having to throw across his body with a batter in his view. Not sure I've ever seen a breakdown of a runner's success stealing 2nd compared to 3rd. I would also mention that Frenchy is only one for five in steals in the month of August, so he may be dinged up and need to curb the aggression a little. On the other hand, Ubaldo Jimenez is a nasty pitcher who was "on" last night, so sometimes you have to try to steal a run from a pitcher like that with three rookies coming up.
OK, I agree with everyone that says it's more complicated than it might first appear. Kind of. The numbers I used are totals overall and (as I said) I don't have the matchup breakdowns.
So maybe the number is higher than 70% (and Jim is right, most base stealers think third is easier than second) or it might be lower than 70%.
What I can say with assurance is the Royals believed the odds of stealing and then having Giavotella put the ball in play were higher than letting the next two hitters swing away or Frenchy wouldn't have been running.
These guys aren't dopes and they've got all the numbers broken down for every situation.
And, Curtis, like Jim I don't have total faith in Fangraph's numbers. Sounds like the don't take account of what base stealing and aggressive base running does for the hitters at the plate (more fastballs, defense out of position, rushed throws, distracted pitchers, etc.). But you're right, Stephen Maturin is one of the all-time great characters in literature: a deadly killer who has a hard time climbing on board a ship without drowning himself.
And, Roger, I never said you would steal in that situation every time. What I did say was in that particular situation it was clear-cut: Frenchy stealing was their best chance.
And Ned Yost apparently agreed with me. (Although he's the same guy that brought in Collins with the bases loaded so maybe that's nost such a hot testimonial.)
On the other hand, look at all the fun we're having arguing.
Lee, did you read the Bill James Baseball Abstracts back in the day? I loved those things. The Elias Baseball Analyst was almost as good. Everything now is based on fantasy ball. It's amazing to think how much Bill James has changed the game.
Since we are discussing the running game, let's talk catching. As word travels around the league about Perez's arm and guys experience his snap throw to first, it's natural that they would shorten their lead by maybe a half step. Doesn't seem like much, but next time there is a play at the plate, imagine that runner a half step closer to the plate. He'd probably be safe. Combine that with a pitchers confidence in him to call the game, or to stop a ball in the dirt with a runner on third, and you have a guy who is helping win games even if he goes hitless.
Lee, I'm sure you went over this before, but where did you grow up? What team and players did you like? What made you fall in love with the game?
About the Frenchy steal attempt odds:
The easiest way to figure out whether it was a good idea to attempt a steal is to look at how many runs the team is expected to score if he doesn't attempt a steal and compare that to how many runs the team is expected to score if he does attempt a steal.
If Frenchy doesn't attempt a steal, the Royals have a runner at 2nd base with 1 out. From 1993-2010, that situation led to on average .721 runs. See: http://www.tangotiger.net/re24.html
If Frenchy does attempt a steal, the play will either result in a runner at third with 1 out (.989 expected runs) or the bases empty with 2 outs (.112 runs). If we say Francoeur steals successfully on 70% of his attempts, we expect .726 runs in the inning if he runs.
The expected runs for those situations are averages, and the situation with Gio and Perez coming up might differ from the 1993-2010 MLB average (Gio and Perez might not be as good at driving runners in from second, but Gio might also not be as good at hitting sac flies), but the MLB average is a good starting point. And it turns out that attempting a steal at a 70% success rate in that situation nets you an increase of .005 expected runs in the inning. In other words, it almost exactly a wash.
Of course, I don't think Frenchy really had a 70% chance of successfully stealing that base. He had a streak of success early when he was catching teams off guard, but he's been much worse than that lately. Anything less than a 70% success rate costs the teams runs in that situation.
"From 1993-2010, that situation led to on average .721 runs."
Unfortunately, the craving for large sample sizes gets us so far away from the game reality as to lose usefulness. In last night's situation we had Ubaldo on the mound and he was dealing, so that is nowhere near an average pitcher as in Tom Tango's study. We also had Johnny Giavotella at the plate and Sal Perez on-deck, two rookies with only a few weeks major league experience, so the batters were below what would have been Tango's average.
I do agree that Frency's odds on that play may have been below 70%. He's had a bad August, 1-4 on the bases, and was 5-3 in July, so the league may be getting him booked, he may be dinged and lost a step, or he may be running more based on situations demanding an extra base with Gio, Moose, and Perez following him. His relatively weak July and August correspond to his drop in the batting order to fifth where he no longer has Billy following him.
Good points, but I still agree with sending him. I just don't have enough faith in Gio and Perez yet against a dominant pitcher to bet on 26% or less in a tight game.
In attempting the steal, you're still counting on Gio to beat Jimenez with a hit, a deep fly, or a ground ball that's hit a bit away from an infielder.
The run expectancy in this situation is certainly lower than it is for the 1993-2010 averages, but it's lower for both the scenario where Frenchy doesn't attempt and for the scenarios where he does attempt. The cutoff for sending Frenchy in that situation really was about 70%.
"In attempting the steal, you're still counting on Gio to beat Jimenez with a hit, a deep fly, or a ground ball that's hit a bit away from an infielder."
True, but a slap hitter has a decent chance of a ground ball to the right side and a passed ball, wild pitch, balk, or error also score the runner on third. Neither view is a slam dunk, but I prefer the aggressive rather than passive play. This time it just didn't work.
Gene: You're right about runners shortening leads with Perez behind the plate, that will have an affect on runners going first to third, second to home, breaking up double plays, and so on.
You're also right about Bill James, but you may want to check out a guy named Earnshaw Cook, who predates Bill James. I don't if Cook deserves credit for starting the sabermetrics craze, but he was one of the earliest people to statistically analyze baseball in a thorough way. He also got a lot of stuff wrong, but he may have started people thinking about baseball in new ways.
As for my background: grew up near Sacramento,California, followed the Giants and later the As, moved to San Diego for two years, moved to KC and adopted the Royals, but was never a serious student of the game until I met Clint Hurdle in 1990.
He was managing a AA team for the Mets at that point and if you wanted to talk about the game with him, you had to pay attention.
Eric: I've looked at those run expectancy charts and, once again, I wouldn't accept them as gospel. I've got a posting I've been working on explaining why.
And Jim is correct, everything depends on who is actually playing, not an overall average of all teams over seven years.
Ok, let's play it simple and just go by batting averages of the players involved.
Gio bats .265 and Perez bats .245.
So the chances of Gio or Perez getting a hit (a single from either likely scores Francoeur) is equal to 1 - ((1-.265)*(1-.245)) = 45%.
So if Frenchy doesn't attempt a steal, he scores 45% of the time.
Assuming a 70% success rate, if Frenchy does attempt a steal, he's out 30% of the time, and if he's successful, he needs Gio to put the ball in play somewhere good or he needs a Perez hit.
Let's say between strikeouts, lineouts, pop flies and hard ground balls right to infielders, Gio drives in the runner from third 60% of the time. If Gio fails, then Perez needs a hit to score the runner. So the odds of Frenchy scoring on a steal attempt is equal to .7(1-(1-.6)(1-.245)) = 48.9%.
That gives you 45% vs 48.9%, and remember that the steal attempt also reduces the chance of a 2+ run inning because if Frenchy doesn't risk an out on the basepaths, there's a better chance for Gio to score in the inning.
Eric: I put a response on here and then realized it was probably incorrect (it was), so I called a buddy who taught statistics and probability.
He figured letting the guys hit away at about the same rate of success you did, 45%, but he also figured the steal and putting the ball in play at a higher number (although that was my fault, if the infield's back there only three places the ball could go that wouldn't score Frenchy, pitcher first or third depending on their depth, so we used a higher number than you did).
Either way, both of you know statistics and probablility better than I do, but we're all ignoring reality to come up with those numbers.
Frenchy's rate of success stealing bases, Giavotella's chance of putting the ball in play or getting a hit and Perez's batting average could've been higher or lower than usual based on the people, situation or conditions involved.