Jagr carries the puck across the line nearly 3/4 of the time...like an oldtimer.

Jagr prefers the carry-in to the dump-in, as most old-timers would.

If you’ve even been to a “Legends of Hockey”-style old-timers game, you may have noticed a certain, distinct style of play from the old ex-NHLers. It stems from what I believe they would describe as “a general disdain for forechecking,” if you caught them to talk about it before their post-game beers so they could still capably use words like “disdain.” They don’t find working hard to retrieve something they already have appealing. Who does, really?

When they push it up on the rush and find that the d-men are pressuring them before the blueline, they swing it back to their defensemen. They loop back then forward again, looking for the eye of a needle to thread a puck through, don’t see it, then regroup again (usually just the puck regroups, they tend to be too lazy to come all the way back). By then the other team is chasing out of position, the next rush finds a seam, and the puck’s in the back of the net. They’re almost soccer-esque in their attack. Games end 9-2 in favour of 55-year-old ex-pros against 25-year-old firefighters because they control the puck, pick their spots, and don’t dump it in.

As a general rule in hockey, dumping the puck in is never ideal. The goal of the dump is to get the puck in your opponent’s zone then regain possession (while avoiding a turnover), so if you can cut out a few of those steps and just have possession in your opponent’s zone, you would (plus you have it while attacking on carry-entries, as opposed to cycling). The problem is, NHL players are too good to just allow you easy zone access when they have numbers back, and they’re too good to simply run the puck back to your d-men after a stymied rush so you can swing up again like the old-timers. You almost never want the puck moving backwards in the pro hockey game, so when you approach the opposing blue and get cut off, your ideal situation is the soft chip or the hard wrap with puck support – you want to put the puck in a place where your team has the best chance of retrieving it. You don’t want to have to navigate the neutral zone minefield a second time.

Over the past few years hockey’s advanced stat community has done some research on zone entries, specifically the excellent Eric Tulsky of Broad Street Hockey and NHL Numbers, who presented this paper at the MIT Sloan Sports Analytics Conference. I’m not great with numbers, but I’m trying to keep up because I’m terrified of becoming the old guy yelling about plus-minus while the majority of people inside and outside hockey have long since dismissed the weight it used to carry. Anyway, I recognize from a player’s angle, just as Eric did as an analyst, that entering the offensive zone with puck possession gives you a lot better chance of scoring than the dump. Numbers, eyes, common sense, same page.

Here’s the chart from his Sloan paper that shows the data he came up with after a year of research. They charted the Flyers/Wild for a full season, the Caps and Sabres for half-seasons, and the rest of the league for 7-10 games each.

carries vs. dumps

In a nutshell, carrying the puck in results in roughly twice the goals of a dump-in.

Since this research was done and made public I’ve been seeing some comments come up in my Twitter feed about zone entries and player evaluation, so I thought I’d weigh in on what information I think can be gleaned from tracking it. Eric T. also wrote this up explaining what he thinks can be gained from having this data (settle for dumps less) . We’re pretty much on the same page, I’m just going pure eyeball/hockey-experience versus a season of data.



I think tracking zone entries would provide value to a team as a whole on the defensive side of the puck, and even more specifically for whichever coach is running the d-corps. If you see from the data that you’re allowing a disproportionate amount of carries into the zone versus dumps, you can make strategic adjustments to force more shoot-ins. (The vast majority of neutral zone strategy on the penalty kill is designed to do just that, while the hope 5-on-5 is to force a decision before the red line — your ultimate goal is to cause a turnover or icing there.)


On the offensive side of the puck, I think it’s kinda cool that the numbers can flesh out differences between players. Skill players are more likely to try to put the puck under defenders sticks (gasp, a risk!) in hopes of getting the puck across the blue with possession (as in, they carry it in more, take more chances). Buuut, they have a lot of leeway because they provide the team’s offense and sometimes their risky behaviour turns into goals for. Third and fourth liners get less minutes, so turnovers are more glaring for them, and errors tend to cost them even more Time-On-Ice. That combined with a fourth-liner’s role literally being don’t get scored on, and a dump makes sense for those low-liners.

Knowing that stuff, if a coach were able to see that a third-liner still had a good percentage of successful carries vs. dumps, he might be able to reason that he’s more skilled than was previously believed, and maybe loosen the reins on him a bit. As a player, attempting more carries is always a game of risk/reward.

Free will is limited

For the bulk of possessions on the offensive side of the puck, dump-ins versus carries are out of your control. If you’re forced into a situation that involves burning a guy 1-on-1, which has a success rate of about 2% at the top levels, it’s undeniably smarter to get the puck in deep instead of risking a turnover. And, even the biggest meathead in the league won’t dump the puck in if he’s got open ice in front of him, so I don’t think you can really tell all your forwards to take more risks at the blueline.

Eric T. actually wrote about this idea, speculating that the risk involved in the turnover would be out-weighed by the amount of good that could come from a zone gained, but I’m not entirely sure that’s true if Ben Eager is gambling more at the blue (despite what the numbers may currently say about equal effectiveness of a carry-in for all, which seems off). Those fringe situations are awfully tough to create out of, and unless you’re talking top line guys I just don’t want to tell Matt Martin to give ‘er a go at the blue more often. I’m not ready to assume you’ll see team gains.

The wrap-up

If I’m taking anything from the information Eric and crew have dug up, it’s that I might tell my top line forwards to try to dump less, and I might encourage my defenseman to pressure harder at the blue lines. That’s at least a start, as the data shows possession in those areas truly does make a difference. You risk getting burned if your gap control at the blue is poor, but given that twice as many goals are scored on possession entries than dump entries, it might just be worth the odd roasting.

I don’t yet think zone entries are particularly valuable when assessing a single player just yet. In making player-by-player comparisons on a single team, maybe, but given that so many – SO many – decisions are dictated by the opposing team, I think those numbers might be more circumstantial than reflective of individual play.

Eric says the numbers show that coaches in most sports are too conservative in general (going for first downs on close fourth downs should happen more, for example). Hockey coaches do frequently give their pre-game speeches and draw lines five feet on either side of both blue lines, and basically say “these are the most important areas of the ice for us. We do not turn the puck over here. We get close pucks out, we get close pucks in.” Safety first is the motto, and maybe it shouldn’t be.

But for now, I can only see zone entries as a stat useful in the big picture (which can obviously change down the road). They might be able to help a team highlight a defensive weakness, they can help define certain player styles, but I’m still not sure I see a solid way to use them beyond that just yet.

Comments (18)

  1. Wondering about the delineation of a carry-in vs. dump in? If a forward carry’s the puck in the zone, is forced to either stop or delay just inside the blueline (in the first 5 feet) and then gets it deep, is that a dump or a carry? I would think it’s probably classified as a carry, but from the perspective of scoring a goal, I would think it has odds more like those of a dump from just outside the blueline, you’ve simply gained an extra 5 feet before dumping.

    • I believe, in terms of this particular set of metrics, that if the puck carrier enters the zone with the puck in his possession, that is considered a carry in. The defining moment for either a carry in or dump in is the blue line. Unless someone is more familiar with this than I am and could provide clarification, I would surmise that it would be incredibly difficult to establish solid data if you turned your constants into variables. One could argue for different parameters, but the data seems to reflect a pretty significant trend as it stands. However, I’m sure we will see improvements to all advanced hockey metrics as they become more mainstream.

  2. “I don’t yet think zone entries are particularly valuable when assessing a single player just yet. In making player-by-player comparisons on a single team, maybe, but given that so many – SO many – decisions are dictated by the opposing team, I think those numbers might be more circumstantial than reflective of individual play.”

    Well put, Jason. I think this is one of the real challenges in analyzing positional sports. Even defensive metrics in baseball where you are more or less standing in a spot based on hitting charts, and performing a rote set of movements, even there it’s hard to make heads or tails of the data. I like the new analysis of hockey through numbers, but i feel that it does more for me to highlight how the really ‘great’ players aren’t simply robots that play the percentages, as much as people whose bodies are better wired to understand and react to what’s happening around them from moment to moment.

  3. The only way they might be really useful is the same way PDO is useful – to find the huge outliers, for teams and players. If there are way too many dump ins, obviously those aren’t all being forced.

    Good stuff though Justin.

    • Why is that obvious? Defensive oriented players aren’t going to get as many stretch passes. I think it’s hard to assume that players are going to have similar opportunities to dump v. carry.

  4. Maybe this gets mixed with a new stat called, “dumped it in and recovered” vs “dumped it in and other team recovered, executed breakout”. Sometimes the dump is just to wear down the other guy a little, kill a bit of clock, and then line change. If a line dumps it to the corner, bangs around a bit, and then skates out to a line change, is that a bad strategy in the grand scheme of the game? Sure, a goal would be better, but no goal against is pretty good too.

    I know “goals” and “shots” are the gold cup of stats right now, but I feel like this concept needs fleshed out in more detail still.

  5. While I think it provides some interesting insight, I have the usual problems with this stat as many of the advanced stats. It makes some pretty big assumptions, the biggest one being that correlation = causation (it doesn’t). It simplifies the game so much too, which I suppose is the purpose of stats, but it still bugs me. If a player has a breakaway, then that counts as a carry in….but can we really compare that to a dump-in in terms of goals scored?

    • If you’re getting enough breakaways for it to be statiscally significant against your dump-ins, you are doing something very, very right.

      • Yes, but a 4th liner might dump it in every time, except for breakaways and odd man rushes.

        In other words, the QUALITY of carry-ins is hard to measure, since the situation differs in each case, and different players respond differently to each on-ice situation.

  6. In the opening paragraph you meant pre-game beers when talking about the old pro’s…….. right?

  7. my best puck handler ask about dump-ins last night before practice. This is the kid that goes end-to-end almost every time he has the puck. So i asked him”you just busted your butt to get the puck, now you want to give it away and go get it again or would you rather play keep away?” I don’t have the stats but I would guess upwards of 80% of ES goals are scored on the rush within 7 seconds(1st or 2nd shot attempt) of crossing the blue line. If you are trying to outscore the other team-then trying to enter the zone with possession and support(hey guys without the puck your important too) is the play to try for first and the dump is a last resort.

    • Mike, 62% of all even strength goals (give or take literally 1%) are scored within the first 10 seconds of the puck crossing the blue line OR the attacking team gaining possession of the the puck in the offensive zone (i.e. A defensive turnover).

      This holds true in regular season AND play-offs… At least over the last 2 seasons, anyway.

  8. my guess was off but on the right track. Thanks

  9. Admit I haven’t read the paper, but how did they control for the fact that different players carry the puck in vs. dump it based on skill? Also how did they deal with when in the shift the zone entry happened (fresh players carry, tired ones dump it)?

  10. If your goal is to make hockey sound as boring as baseball, then bravo, nerds, you’re succeeding.

    • Because using the information available to better understand the game is useless, right? I’m very sure that actual NHL front offices using analytics at least to some degree to make their decisions (just like in baseball). Why are some fans so scared of numbers?

  11. Way late to the party, but glad I found this article anyways!

    I think the power of these advanced stats is their ability to compliment video analysis of the game. If I were a coach of a team and I saw the results of this study, I’d want to know the stats for my team. You’d expect first liners to have a higher percentage of carry-ins vs dump ins than 4th liners, but if you saw some player that was an outlier, it would suggest to look closely at the bulk of their zone entries through video. If the player had a much higher rate of dump ins, maybe they are making poor decisions and dumping the puck in when they had space to carry it in. If a player had a much higher percentage of carry ins, you would want to see if it was entirely because they are more skilled, or if they are doing something slightly different tactically that might be applied to everyone on the team.

    If your team as a whole has a dump in ratio that is higher than average, you would want to see if its decision making or tactics that are causing this and then try to do something about it.

    What the people who oppose advanced stats seem to miss is that advanced stats aren’t to replace watching the game, but to compliment it. Good advanced stats won’t tell you everything, but they will act as a filter to tell you what is important to study, and what is noise and would be a waste of your time to study.

Leave a Reply

Your email address will not be published. Required fields are marked *