Doubling up
Posted by Andy on April 19, 2010
Since the 2000 season, there have been 99 occasions in which a single player has hit at least 2 triples in a game. Here are the instances since 2009:
| Rk | Player | Date | Tm | Opp | Rslt | PA | AB | R | H | 2B | 3B | HR | RBI | BB | IBB | SO | HBP | SH | SF | ROE | GDP | SB | CS | WPA | RE24 | BOP | Pos. Summary | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Curtis Granderson | 2010-04-15 | NYY | LAA | W 6-2 | 4 | 4 | 1 | 2 | 0 | 2 | 0 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0.099 | 1.217 | .680 | 8 | CF |
| 2 | Erick Aybar | 2009-09-06 | LAA | KCR | W 7-2 | 5 | 5 | 1 | 3 | 0 | 2 | 0 | 1 | 0 | 0 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0.128 | 0.939 | .443 | 2 | SS |
| 3 | Dexter Fowler | 2009-08-22 | COL | SFG | W 14-11 | 5 | 3 | 2 | 2 | 0 | 2 | 0 | 2 | 2 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0.210 | 3.215 | .830 | 2 | CF |
| 4 | Cory Sullivan | 2009-08-12 | NYM | ARI | W 6-4 | 5 | 4 | 0 | 2 | 0 | 2 | 0 | 1 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0.143 | 1.152 | .810 | 1 | CF |
| 5 | Jacoby Ellsbury | 2009-06-23 | BOS | WSN | W 11-3 | 5 | 4 | 1 | 4 | 0 | 2 | 0 | 3 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0 | 0.282 | 4.075 | 1.005 | 7 | CF |
| 6 | Andrew McCutchen | 2009-06-08 | PIT | ATL | L 6-7 | 7 | 7 | 2 | 4 | 1 | 2 | 0 | 1 | 0 | 0 | 2 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0.172 | 1.746 | 1.781 | 1 | CF |
| 7 | John Buck | 2009-04-30 | KCR | TOR | W 8-6 | 4 | 3 | 0 | 3 | 1 | 2 | 0 | 5 | 0 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 | 0.224 | 4.022 | .903 | 8 | C |
| 8 | Jeff Francoeur | 2009-04-12 | ATL | WSN | W 8-5 | 4 | 4 | 2 | 2 | 0 | 2 | 0 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0.095 | 1.328 | .717 | 6 | RF |
Over that time time period, there have been 9,110 instances in which a player hit one triple. (In this case, individual games in which more than 1 player hit a triple are counted more than once in this list.)
That's a ratio of 9,110:99 for number of 2+ triples games to 1 triple games. That ratio simplifies down to 92:1.
I wonder how this compares to other types of hits.
There have been 168,359 games in which a player has gotten one single (and not more than 1 single.) Again, any games featuring more than one such performance are counted multiple times.
There have been 75,717 games in which a player has gotten one double.
There have been 46,297 games in which a player has gotten one home run.
For 2 or more of each hit type, the numbers shake out to:
singles: 57,890
doubles: 6,866
homers: 3,055
So we can calculate the ratios for each type of hit. These, then, are ratios of two or more hits of a particular kind to just one hit of that same kind.
singles: 2.9:1
doubles: 11.0:1
triples: 92.01:1
homers: 15.2:1
These ratios should be, pretty much, in line with the general odds of getting any one type of hit. After all, they are pretty much independent events. What I mean by this is that just because a guy got a single in his first at-bat, that doesn't mean he's any more or less likely to get a single in his next at-bat than he would be if he made an out the first time (or homered.) When I say "pretty much independent" I do leave open a few very minor issues such as players on a hot streak, ballparks that might have a greater tendency to yield a certain type of hit, etc.
What, then is the point of this post? There isn't one. I was just curious.
April 19th, 2010 at 2:57 pm
How about a step further?
Double-Doubles (at least 2 of each in one game since 2000)
1B+2B:455
1B+3B: 11
1B+HR:161
2B+3B: 1 (Carl Crawford on 08-02-05) (9 times total in all the PI years)
2B+HR: 32
3B+HR: 1 (Dmitri Young on 05-06-03)(In the years currently covered by PI, there are only 4 such games. The others were Willie Mays in '58, Lew Fonseca in '29 and Lou Gehrig in '28)
April 19th, 2010 at 5:43 pm
Andy-
A bit of a quibble with how "independent" the events are. The frequency of a given outcome is league wide or, really, historic. But that does not mean it provides the odds for a given player. The likelihood we'll see one of these events happen at any given time follows the ratios given. But, for a given player, they can vary wildly.
Albert Pujols hitting 2 HRs in a game is not the same as Rey Ordonez. Yes, each at bat is 'independent' of the previous one. But for Pujols, it's like flipping a coin where one outcome is HR and the other one is everything else and for Ordonez it's more like rolling a die where one outcome is HR and the other five are everything else (obviously, this is grossly oversimplified and not scaled properly). Now, this party is pretty obvious.
But taking it a step further, the guy more likely to hit 2 HRs in a game is also the guy more likely to hit 1 HR in a game. So, if we know a guy already hit 1 HR in a game, we can make a certain (albeit very minor) assumption about his ability to hit HRs. And can assume that he is more likely to hit 2 HRs in the game than the guy who has zero, and not ONLY because he has a head start.
This is like the, "Are you more likely to score multiple runs in an inning with a lead-off HR or a lead-off walk?" Despite what Joe Morgan thinks, you have a far better chance with the former and not only because you are already guaranteed of one run... but also because it can be reasonably assumed you are more likely to score runs off a pitcher giving up HRs than one giving up runs. The difference may be marginal and even negligible, but it's certainly not zero.
April 20th, 2010 at 8:43 am
I think Granderson had another of those 99 instances in 2007. I seem to recall a 2+ Triple game from him that year (he hit 23 in total).
April 20th, 2010 at 8:45 am
Update: Granderson did it twice in 2007.
April 20th, 2010 at 9:13 am
Good points, BSK. Andy, you said "These ratios should be, pretty much, in line with the general odds of getting any one type of hit". Are they? It would be interesting to see if the effect BSK describes is observed or if it all averages out (Pujols' high probability offset by Willy Taveras' low one). In the meantime, I'll take a look at just 2009.
For 2009, the rates were:
1B: 1 per 6.50 PA
2B: 1 per 21.4 PA
3B: 1 per 197 PA
HR: 1 per 37.1 PA
This leads me to believe that BSK is right. These ratios (from 2009) are roughly half of the ones from Andy's set (from 2000-2009). Though I'd like to see the ratios from the full decade for an apples-to-apples comparison, it seems pretty conclusive that a player who's gotten a hit is actually more likely than average (perhaps significantly so) to get another.
April 20th, 2010 at 12:32 pm
To give proper idiotic credit here, it was Tim McCarver who thought that about lead-off walks.
April 20th, 2010 at 1:21 pm
Adding another layer; I would think that a big time HR hitter like Pujols would be about as likely to hit 2 in one game as a marginal HR hitter (say a 20-30/yr guy). The reasoning being that once a guy like Pujols hits that first HR, pitchers are more likely to pitch around him the rest of the game. After a marginal HR hitter hits one, the pitches he sees the rest of the game will remain about the same. That might even out the odds increase you get from being more likely because you have more total HRs, but I don't have data to back up this thinking.
2009 stats for four names chosen (somewhat) at random:
Mark Reynolds: 44 HR, 3 games with multiple HRs
Curtis Granderson: 30 HR, 3 games with multiple HRs
Albert Pujols: 47 HR, 10 games with multiple HRs
Alphonso Soriano: 20 HR, 1 game with multiples HRs
So at a quick glance, it's looking like I'm wrong
April 20th, 2010 at 1:25 pm
Yeah I think the arguments in #2 and #5 are not quite right because the frequency of home runs already takes into account the total number of home runs hit. If Pujols himself is going to hit 30 more this season, the events are still fairly independent in terms of whether they will come in separate games or multiples in the same game (again ignoring factors like his comfort against certain pitchers, etc.)
It's true that it's much more likely for Curtis Granderson to hit 2 triples in a game than, say, David Ortiz, but what I measured here is just the ratio of 1-triple games to 2-triple games. I don't think the results are any different than if I measure it for games where ANY players hit 1 triple or 2 triples. I will check that.
April 20th, 2010 at 1:36 pm
Andy: My comment was based on your statement "These ratios should be, pretty much, in line with the general odds of getting any one type of hit". I was just saying that it didn't look like that was the case.
April 20th, 2010 at 1:41 pm
My statement was an oversimplification, and what you calculated was not what I meant.
What I really meant was--if one went through and calculated it, I think that the general odds of getting a second single in a game, given that a first one has already been hit, is basically the same as getting the first one, once correcting for the number of plate appearances remaining in the game. This is what I mean. Take an average .280 hitter. Let's say that guy get a single in his first plate appearance. I don't think the odds that he gets a hit in his next few appearances goes up dramatically just because he got a hit in the first plate appearance. His odds are still about .280 in each of his subsequent plate appearances, obviously ignoring issues such as specific pitcher matchups, runners on base, etc.