clock menu more-arrow no yes mobile

Filed under:

On the Travis Hamonic-Andrew MacDonald Split, and Stat Analytics

The pair reshuffling experiment creates new questions. The fun of sports is how elusive the answers can be.

Your fault, my fault, let's blame the goalie.
Your fault, my fault, let's blame the goalie.
Jana Chytilova/Freestyle Photo

If you perused our last two posts (and comments) or more conventionally, just looked at the original source, you already know Arthur Staple's interesting tidbit on the training camp breakup of the Travis Hamonic-Andrew MacDonald pairing. From Staple's a.m. blog post at Newsday:

For those fans interested (as I am) in advanced stats, the Isles are using them to assess defensive pairings, and they came away with the idea of splitting Andrew MacDonald and Hamonic. Neither defenseman seems terribly excited by the idea -- he never said so, but I presume MacDonald’s being against it partially stems from the fact that this is a contract year for him, so the more quality minutes for him the better -- but we’ll see if analytics pay off.

We covered possible reasons for this experimentation a few days ago, but the "advanced stats idea" makes for an interesting one and reminds us of garik16's post earlier this summer, which looked at the pairing via the neutral zone entry perspective. You may have missed it while on the beach, or while watching bad baseball (a redundant term), but his conclusions were provocative.

Maybe We Only Think They're 'Top Pair' Because That's Their Only Role

Though we traditionally see this pair as a strong "shutdown" pair, that may be by default and assumption rather than actual evidence. In a word, garik concluded from last season's numbers:

They didn't do a particularly good job limiting top opponents. In fact, Opponents had a better offensive performance against Hamonic and AMac than they did on average!

So this tale is interesting from that perspective alone: Reputation maybe not matching reality, as well as the whole analytics/advanced stats/underlying numbers/[choose your favorite term here] narrative where teams may be looking behind the boxscore but by their secretive nature will not share much.

On that note, we don't know which stats the Isles are using, nor in what way. When pressed, Staple replied on Twitter:

Contracts? Contracts.

The other interesting aspect here relates to what we touched on in the earlier training camp post, namely: The Isles just gave Hamonic a long-term contract, while MacDonald is entering his walk year. So who is the better spouse in this pair? In July Garik16 posited:

Well it's not clear. For one, this is Hamonic's first year like this....but it's NOT AMac's first year like this - AMac had some meh #s last year, while Hamonic's #s last year were decent (and better than AMac's for whatever reason) if not great. On the other hand, Hamonic's best possession #s came in his rookie year.

Money talks, or something. The Isles have made a clear and major commitment to Hamonic, while GM Garth Snow has said there have been initial talks about MacDonald's next contract. Clearly the Isles like him. But just as clearly, they have a lot of prospects in the system who will gun for blueline jobs. Whether it's AMac walking or being traded, or another D-man eventually leaving (Matt Carkner? Two more years. Brian Strait? Three.), new blood will come in and that will at least partially hang over AMac contract negotiations.

Players Don't Read Microstats

For our own dramatic interests, there is the evergreen issue of players being brought up to play and see the game a certain way. They are human, they feed off relationships, and even more than old-school baseball managers they may be the last to embrace an analysis that comes not from "feel" and "eyes" but from reviewing aggregate data over way more repetitions than any human memory can meaningfully recall.

To that end, Staple's bit about neither Hamonic or MacDonald sounding "terribly excited" about the split adds an interesting footnote. Both will have to believe in it for it to work. Isles PR and news coverage of the team has praised the pair's top-competition role the last few they're supposed to think an alternative combo is better? It will take some cognitive work. (On that note, this story came up in comments on how baseball's Pirates altered their approach based on data...which some players embraced and at least one did not.)

Who Is The Devil We Don't Know?

Finally, there is one last annoying matter, something which training camp can help but can hardly conclude with any authority (preseason isn't informative nor long enough to test thoroughly): Just because you break up an underperforming (relative to role) pair doesn't mean you have a better alternative at the ready. And the lovely sports dilemma is you won't know more about the alternatives until you've given them a chance to hurt you.

For now, it's Hamonic with Brian Strait and MacDonald with a mix of candidates, including prospect Matt Donovan. Will a reshuffle yield actual improvement? And if so, in what combination?

The elusiveness of the answer is one of the many reasons we watch* these silly games.

(*P.S.: "Watch the games"...? In an analytics post? Sacrilege.)