That is the question. My previous comments on the futility of the sacrifice bunt provoked some discussion. While it appears that Melvin actually uses the bunt less often than most National League players, I thought it might be enlightening to look into the specific times when he did so.
Before going into that, worth pointing out an article on BaseballAnalysts.com which analyzes bunting in more detail. It uses the run expectancy tables - which give you the expected number of runs with X men on base and Y outs - but breaks them down further by batting order. This makes a big difference. In the NL, with your leadoff hitter up to start an innings, the expected number of runs is .542; but if your #7 is up, it plummets one-third, to only .363. One of the conclusions there is interesting: "when teams actually bunt they sometimes do, in fact, increase the expected runs over the remainder of the inning, particularly late in the order with no outs and a runner on second or first and second."
The article is also very useful, in that it includes charts which provide the chances of scoring at least one run - often the aim of a bunt, when you're tied or behind by one. This does alter the bunting landscape: for example, a runner on second with no outs will score that runner 60.9% of the time, overall. But a successful bunt moves him over to third with one out - from where he'll score 64.8% of the time. Again the exact percentages vary depending on where you are in the order, but it seems my original evaluation - which, er, I believe was, "Bunting is bad, m'kay" - may have been overly simplistic.
Looking at Arizona last season, there were a total of 21 occasions in 2006 when a non-pitcher was credited with a sacrifice, divided up among eleven players. Orlando Hudson led the way with four, while Damion Easley and - perhaps surprisingly - Andy Green, came second on three. But even Chad Tracy and Johnny Estrada had one each. Melvin seemed significantly more bunt-happy later in the season: he had eight in August, compared to only nine in total before the All-Star break. However, three of those August bunts came in one game.
Not all of these may necessarily have been ordered by Melvin; some may have been attempts to bunt for a hit with a runner aboard. But with that caveat, let's take a look at them. For each event, I've logged the date, score and circumstances - along with whether I think it was a good move, and the final score.
Hudson, April 25 vs. San Diego
Score: 1-0, top 3rd
Situation: 2nd, no outs.
On-deck: Tracy and Gonzalez
Analysis: The run scored on a Jackson single, but we were facing Jake Peavy, and so runs could be considered at a premium. Very early on, but Hudson was struggling (hitting .243) AGREE
Final score: AZ 7, Padres 0
A.Green, May 2 vs. Los Angeles
Score: 7-7, bottom 7th
Situation: 1st/2nd, no outs
On-deck: Snyder and pitcher's spot
Analysis: Not a lot of pop coming up, late in a tied game. Good call to get the go-ahead run to third, stay out of the double-play. Plenty of bench left for the pitcher. AGREE
Final score: AZ 10, Dodgers 8
Easley, May 7 vs. Cincinnati
Score: 7-9, bottom 9th
Situation: 1st/2nd, no outs
On-deck: Green and Davanon
Analysis: It did get the tying run into scoring position, and DaVanon then singled. However, Estrada lined into a double-play at first. That's not Melvin's fault. AGREE
Final score: AZ 8, Reds 9
Easley, May 14, vs. St. Louis
Score: 3-3, top 7th
Situation: 1st/2nd, no outs
On-deck: Tracy and Gonzalez
Analysis: Meant a sac fly would give us the lead, and Tracy couldn't GIDP. He did neither, even though the Cards went to their bullpen, and doubled home both men. AGREE
Final score: AZ 7, Cardinals 6
Counsell, May 22 vs. Pittsburgh
Score: 0-1, bottom 7th
Situation: 1st, no outs
On-deck: Byrnes and Gonzalez
Analysis: We manufactured the tying run: after the bunt, a groundout, and a bunt single by Gonzo tied the game. If that was all planned, it was nicely done by Melvin. AGREE
Final score: AZ 4, Pirates 3
Estrada, May 22 vs. Pittsburgh
Score: 1-1, bottom 8th
Situation: 1st/2nd, no outs.
On-deck: Hudson and Clark
Analysis: The runs did score, but as the result of an E9 on Hudson. Clark then K'd, which would have ended the inning. Still, late in a tied game, getting the go-ahead run is what matters. AGREE
Final score: AZ 4, Pirates 3
Easley, May 27 vs. Cincinnati
Score: 3-0, top 7th
Situation: 1st/2nd, no outs.
On-deck: Counsell + Byrnes
Analysis: Easley pinch-hit for Cruz, which seems weird if you're calling for a bunt. Why not use an experienced pitcher for that, rather than wasting a position player. DISAGREE
Final score: AZ 7, Reds 0
A.Green, June 1 vs. Atlanta
Score: 0-0, top 8th
Situation: 1st, no outs
On-deck: Byrnes and Tracy
Analysis: Ramirez had handcuffed us to three hits in seven innings, so playing for one run was okay. Worked nicely, as the pitcher's error put men on 2nd/3rd, still with no outs. AGREE
Final score: AZ 2, Braves 1
Counsell, June 24 vs Anaheim
Score: 4-4, bottom 9th
Situation: 1st, no outs
On-deck: DaVanon and Gonzalez
Analysis: With hindsight, especially, this blew. DaVanon K'd, they walked Gonzo to get to S.Green, who K'd too, stranding the runner. Bunting only makes sense if those following execute. DISAGREE
Final score: AZ 4, Angels 6
A.Green, July 28 vs. Houston
Score: 7-7, top 10th
Situation: 1st, no outs
On-deck: DaVanon and Jackson
Analysis: Green pinch-hit for the pitcher's spot, but the sacrifice was in vain, with the next two grounding out harmlessly. Still had most of our bench left at that point. DISAGREE
Final score: AZ 8, Astros 7
Jackson, August 6 vs. Houston
Score: 2-2, bottom 6th
Situation: 2nd, no outs.
On-deck: Quentin and Tracy
Analysis: The run scored, but probably would have anyway, since Quentin singledy. A tied game, but given the batters coming up, and Jackson's abilities, I'd have let him hit away. DISAGREE
Final score: AZ 4, Astros 3
Byrnes, August 9 vs. Giants
Score: 1-1, bottom 5th
Situation: 2nd, no outs
On-deck: Hudson and Gonzalez
Analysis: Hudson grounded out, but it was all rendered somewhat moot by Gonzo's homer. The runner was Miguel Batista, which merits factoring in and brings me down on the side of AGREE
Final score: AZ 5, Giants 3
Hudson, August 15 vs. Colorado
Score: 1-1, top 18th
Situation: 1st, no outs
On-deck: Gonzalez and Jackson
Analysis: Hard to argue here. Our last reliever had already pitched two innings, so we needed to make the most of this chance. We scored on a single by Gonzo. AGREE
Final score: AZ 2, Rockies 1
Snyder, August 19 vs. San Diego
Score: 4-1, top 9th
Situation: 2nd, no outs
On-deck: Quentin and Byrnes
Analysis: I think sacrifice bunting when you're ahead is almost inexcusable. An insurance run, maybe, but you already have the upper hand: better to go for the jugular. DISAGREE
Final score: AZ 5, Padres 3
Hudson, August 25 vs. Los Angeles
Score: 6-6, bottom 10th
Situation: 1st, no outs
On-deck: Gonzalez and Tracy
Analysis: Tracy GIDP'd, stranding the runner at second, after an intentional walk to Gonzalez. The latter, at least, should have been forseen by Melvin. DISAGREE
Final score: AZ 9, Dodgers 7
Drew, August 25 vs. Los Angeles
Score: 6-6, bottom 11th
Situation: 1st, no outs
On-deck: Quentin and Estrada
Analysis: More by accident than design, this worked out: an error left the runner at third, who then came home on a sacrifice fly. Not much offense coming up by this point. AGREE
Final score: AZ 9, Dodgers 7
Drew, August 25 vs. Los Angeles
Score: 7-7, bottom 13th
Situation: 1st/2nd, no outs
On-deck: Estrada + pitcher's spot.
Analysis: Ick. After the bunt LA, inevitably, walked Estrada, and we had to use Hernandez 2.0 as a pinch-hitter. He fanned, Byrnes did too. A total waste. DISAGREE
Final score: AZ 9, Dodgers 7
Byrnes, August 26 vs. Los Angeles
Score: 2-3, bottom 8th
Situation: 1st, no outs
On-deck: Hudson and Gonzalez
Analysis: Hudson drove the run in with a single. Without the bunt, the runner would likely not have been on second, or have scored that inning. AGREE
Final score: AZ 3, Dodgers 4
Hudson, September 15 vs. Colorado
Score: 0-1, bottom 1st
Situation: 2nd, no outs
On-deck: Gonzalez and Tracy
Analysis: We were facing Kim, a pitcher we positively own [his career ERA vs. AZ is 7.71 in 23.1 IP]. Why play for a single run in the first inning? Very, very strongly DISAGREE
Final score: AZ 5, Rockies 1
Tracy, September 16 vs. Colorado
Score: 6-6, bottom 16th
Situation: 1st, no outs
On-deck: Estrada and pitcher's spot
Analysis: After the bunt, the runner (Young) then stole third, a much riskier play. Why not steal second, then bunt him to third? He scored on a sac fly by PH Snyder, but the tactics here seem bizarre. DISAGREE
Final score: AZ 7, Rockies 6
Drew, September 25 vs. Giants
Score: 4-1, top 9th
Situation: 1st/2nd, no outs
On-deck: Young and Montero
Analysis: Again, shows little confidence in your team, when you bunt with a three-run lead in the last inning. The upcoming hitters both singled, but and seems pointless. DISAGREE
Final score: AZ 7, Giants 1
Looking at the above, it seems that Melvin will only ever bunt with no outs: he didn't ever do it with one down already. Maybe he was aware of the article quoted earlier in the story? Of those 21 instances, 13 were in tied games, with the other eight split equally between the D-backs being ahead and behind - in all of the latter, we had the tying run on base. I agreed with the majority of the decisions, but only just: 11-10. On the other hand, there were only about three or four decisions which I reckon were completely wrong - and that's much less than it seemed.
[As an aside, it seems Melvin isn't entirely a stathead, still believing firmly in the elusive "clutch hitter": "Absolutely," said D-backs manager Bob Melvin, when asked if the clutch hitter really exists. "And sometimes it's not the guy that's hitting .330. Sometimes it's the guy that's hitting .250 that rises to that occasion." Er, which would seem to me to indicate that it's largely a random factor, if any player is capable of 'rising to the occasion'. I do actually believe some players are "clutch", but that the effect on performance is so small, it almost always gets lost in the noise of random variation.]
Obviously, we shouldn't remove the bunt from our armory entirely; if the opposition knows it isn't coming, they can play accordingly. But part of me still does think it's used too much: especially in a good-hitting lineup like (hopefully!) AZ 2007, it should generally be reserved for instances where, for example, the third-baseman is playing back, meaning there's a decent chance of a bunt hit too. It'll be interesting to see how Melvin operates this season, with a significantly different, possibly faster, set of players.
On the other hand, in the 18 different contests where we bunted last year [one game had two, and another three], we went 15-3. Even taking off the four cases where we were ahead at the time, we were 11-3, so I can't complain too much in the bigger picture. Maybe there is some psychological impact on the opposition? Too small a sample to tell though, and I have no interest in looking deeper. Especially not on a Monday. :-)