Who’s better at making defensive pairings: Glen Gulutzan, or dice?

Here’s a fun story from Baseball Between the Numbers:

On Sunday morning, August 13, 1972, the Detroit Tigers had dropped into second place, a game behind the Baltimore Orioles in the AL East. After holding off the Orioles for months, the Tigers were in a free fall with just 3 wins in their last 13 games. Only the Orioles’ mediocre 7-7 record over the previous two weeks was keeping the Tigers from falling further behind.

Billy Martin, in his second year at the helm of the Tigers, decided desperate times called for desperate measures. Instead of the customary impassioned speech to the team or deftly organized “players-only” meeting, Martin decided to shake things up on the field. He wrote down the names of his starters on slips of paper, put them into a hat, and filled out his lineup card by pulling out the names at random. Whether because of the lineup changes or something else, the Tigers won the first game of their doubleheader with the Indians, 3-2. Then, using their traditional lineup for the nightcap, they got crushed, 9-2.

Choosing the lineup is one of the most underappreciated tasks of a sports coach. There’s plenty to consider, and a minor mistake may cost you the game. It’s not as simple as sorting your players from best to worst. Coaches consider things like chemistry, playing style, opposition weakness, fatigue, righty-lefty, etc. It’s not as easy as it sounds.

Billy Martin had his way of doing it, and it worked pretty well for him. Glen Gulutzan prefers his way, and it’s absolutely baffling. For a progressive coach, Gulutzan’s defensive pairings have been a pain for almost every Flames fan. Instead of going with the proven (two-ish years of evidence!) pairings, he likes switching it up. That idea in itself isn’t bad, but he seems committed to his bad pairings regardless of the results. There are much better ways of doing this, but he doesn’t seem to notice them.

Glen has eight choices for six spots, and he seems to pick wrong even with those great odds. It’s gone on too far, Glen. We’re taking away your power to make defensive pairings. In your place, we offer dice.

Rules of the game

  • Billy Martin already did it, and we’ve done it before, so le hat c’est morte. Instead, we have dice.
  • There are two dice, both eight sided. One is for the players, one is for the spot.
  • We roll the player dice first, and the number that shows up corresponds to a specific player. 1 is Giordano, 2 is Brodie, 3 is Hamilton, 4 is Wideman, 5 is Jokipakka, 6 is Engelland, 7 is Kulak, and 8 is Grossmann.
  • Then we roll the second dice, corresponding to spot. 1 is first pairing, left side. 2 is first pairing, right side. 3 is second pairing, left side yadda yadda. 7 and 8 are healthy scratch spots.
  • There’s no objective way to judge these pairings, so we’re going with subjective scales out of 10. There’s a category for players and a category for pairings. Basically, it’s my opinion. Hope you agree with it.
  • The dice get 10 run-throughs. At the end, if they have an individual category score over 50/100, they have won that category. If the dice do 100/200, they have won the game. I was considering raising the standards because we can’t objectively compare the dice’s pairings to Gulutzan’s pairings, but even if some know-nothing dice can dress a 50% optimal lineup, that’s a victory.

Roll 1


Players: Scratching Kulak isn’t that big of a sin, but when it comes at the expense of including Grossmann, it is. I give this a 7/10.

Pairings: Gulutzan’s “good with bad” philosophy goes to its logical extreme. 3/10

Totals: 7/10 players, 3/10 pairings, 10/20 total

Roll 2


Players: It gets even worse. 5/10

Pairings: Also got worse. 2/10

Totals: 12/20, 5/20, 17/40

Roll 3


Players: I don’t think dice can learn, but they appear to be. Still a 7/10, because precedents and whatnot.

Pairings: It still has trouble with the pairings however. Giordano-Grossmann remains a constant on the second pairing, while Kevin and Dennis take the top spot while a more deserving Hamilton-Brodie pairing gets the grunt work. 3/10

Totals: 19/30, 8/30, 27/60

Roll 4


Players: I think the dice have learned the rules of the game and have decided to stick with the 7/10 route for an easy victory.

Pairings: But they seem to forget that the player selection is only 50% of the score. The dice keep getting worse when it comes to actually sorting the players out. 2/10

Totals: 26/40, 10/40, 36/80

Roll 5


Players: Disregard all text above that suggests that dice can learn. They cannot. 1/10

Pairings: Player selection does impact pairings a lot, so it may seem unfair to the dice to judge them because of how they used what was available. However, they can mess up that too. Also a 1/10

Totals: 27/50, 11/50, 38/100

Roll 6


Players: The dice rectified one mistake, but not two. 4/10

Pairings: They’re acceptable, but in reverse order. 3/10

Totals: 31/60, 14/60, 45/120

Roll 7


Players: The dice have excluded Giordano from three consecutive “games.” They did much better when scratching Engelland four times in a row. Take a hint, dice. 3/10

Pairings: Fart sound. 2/10

Totals: 34/70, 16/70, 50/140

Roll 8


Players: Nope, that’s not how you get back into the good books. 3/10

Pairings: I can’t really complain given what the dice selected. If you swap the first and second pairings, it might resemble something Gulutzan would actually put together (provided T.J. Brodie is injured or in exile or something). 5/10

Totals: 37/80, 21/80, 58/160

Roll 9


Players: I know there’s not a lot of love for Wideman left in the Flames fanbase, but surely he deserves it more than Grossmann. 3/10

Pairings: As I said for the previous roll, the fact that this is something Gulutzan would plausibly put together given the circumstances is worrying. 5/10

Totals: 40/90, 26/90, 66/180

Roll 10


Players: I would like to note that Grossmann has only been scratched twice. Mark Giordano has been scratched four times. 4/10

Pairings: These are also probably Gulutzan pairings, except in a terrible order. 2/10

Totals: 44/100, 28/100, 72/200

Final thoughts

Looks like assembling a roster is tougher than it seems! Let’s never complain again.

The dice started off so promising by scratching Engelland four straight times and then dropped off quickly by scratching Giordano four times. It also had nearly no idea where certain players belonged in the lineup, and they got dinged heavily for that. That is okay because they are dice who have no other purpose besides being thrown at a table.

What Billy Martin and his hat proved was not necessarily that conventional management is wrong, just that it isn’t necessarily right. Glen Gulutzan isn’t as dumb as a pair of dice, but he isn’t the best at managing his defence. Lineup decisions are tough, and you can’t do it at random. However, I think that he still doesn’t ice optimal pairings, and it’s obvious to everyone but him. He is a coach at the highest level of hockey with many years of experience. So far in this young season, he broke up the consistent top pairing, gave Nik Grossmann actual ice time, and gave more minutes to Dennis Wideman than Dougie Hamilton.

There is no logic to these moves. Hamilton scored two goals last Friday night and has been one of the better Flames at driving play this season. His reward was going back to third pairing duty.

    • supra steve

      Then Grossmann will play 3 out of every 4 games, and Gio/Brodie/Hamilton will each sit 1 out of every 4 games. So your dice game is a real sound plan.

      You don’t have to like the coach, but the dice game is a little…stupid.

      • JBulls

        Do I really have to explain that I don’t actually want a PAIR OF DICE coaching this team? Guess so.

        You are right that I don’t like the coaching, however. Not sure how anyone can be confident in the coaching of this team right now.

  • Greatsave

    I don’t follow the reasoning of rolling two dice instead of just the one. Surely I don’t think Billy Martin was pulling a name and a position in the batting order out of a hat and matching them together? I would have imagined whoever will pulled out first batted first.

    So wouldn’t it be easier to just roll the one die–a player die–and slot them in from top pairing to bottom to scratches?

    • Really, the reason is that it’s too easy.

      If you already pick the slot and then roll the dice, you only have favourable odds for the dice. I.e, it has eight possible results for the #1 spot, seven for the #2 spot, six for the #3 spot etc.

      If you add a position die, those odds become exponentially worse for the die. There are now eight possible players for eight possible spots, so 64 possible results. Then 49 for the next roll, 36 for the roll after, 25, 16, 9, 4, 1 for the last roll.

      It is much easier to select a spot and then roll for the player, but it’s much funnier when human thought fails against inanimate objects with tremendous odds against them.