Sunday, December 14, 2014

The dump-in as a defensive rather than offensive-play

"Dump and chase" is a term often used to describe teams who employ a system whereby the dump the puck deep into the offensive-zone with the hopes of chasing the puck down, and eventually creating an offensive-opportunity. The truth however, is that the dump-in is a limiting offensive-producer, and is more useful for its defensive rather than its offensive-value.

The reality of the dump and chase is that the a dump-in rarely results in a direct loose-puck recovery; regardless of whether it's a wraparound, soft, or cross-corner dump-in. In fact, according to data I've tracked over the last 4 years, less than 20% of all dump-ins are directly recovered by the forechecking team. In other words, 80% of dump-ins directly result in a loss of possession. By comparison, a player attempting to protect the puck along the boards in the defensive-zone loses possession of the puck only 54.4% of the time.

A more productive way of gaining puck-control in the offensive-zone is by way of loose-puck recoveries following a successful stick or body-check by the forechecking player. On average, this method produces 2.5 times more recoveries of possession than the above-mentioned dump-in LPR.

So why dump the puck in at all?

For its defensive-value.

Using conditional probability we can assign an estimated goal or shot value to any puck-possession event originating from any area of the ice. Expected goal and shot value can be calculated by using the average success-rate of any specific play, as well as the success-rates of all succeeding plays. The goal is to place a value on a play by determining how successful that play has been in the past at either producing a shot-on-net, or a goal.

Estimated goal value is the probability of that play eventually producing a goal, while estimated shot value is the probability of that play eventually producing a shot on net. The top estimated goal value of a play originating in the defensive-zone is is 6.7%. In other words, teams score an average of 6.7% of the time they attempt a stretch pass that produces a breakaway. It's important to understand that this doesn't tell us that team's score 6.7% of the time they attempt a stretch pass from the defensive-zone, it only reflects the EGV of an attempted stretch pass that results in a breakaway. The top EGV of a play originating in the neutral-zone is 10.2%. Similar to the combination expressed earlier, this play involves an undefended controlled offensive-zone entry combined with a shot from the slot.

Simply based on the EGV's expressed above, it is easy to understand the defensive-value of a dump-in compared to a neutral-zone turnover.

Returning to the original point of how a dump-in is a limiting offensive-producer, we see that the estimated goal value of a neutral-zone dump-in is only 0.3%. In other words, on average only 0.3% dump-ins eventually result in a goal, while the EGV's of  a controlled offensive-zone entry, followed by a pass off the rush, and a shot from the slot is actually 5.9%. Again, it's important to remember that this doesn't mean that 5.9% of controlled offensive-zone entries result in a goal, it means that (based on conditionally probability) a goal has been scored 5.9% of the time a play including this combination of events has been attempted in the past.

Although a dump-in will always have limited offensive-value, it's also important to note that any puck recovered following a successful dump-in allows a team to increase the time it spends in the offensive-zone, while limiting the time spent defending in the defensive-zone.

Team's that produce a corsifor% above 50% engage in an average of 14% more events with possession of the puck in the offensive-zone than they spend defending in the defensive-zone. When that difference drops below 12.5% team's CF% generally drop below 50% as well. In other words, using data I've gathered  over the past 4 years teams spend an average of 28.8% of their total even-strength events with possession of the puck in the offensive-zone, while spending an average of 15.7% of their events attempting to remove possession from the opposition in the defensive-zone. If we subtract the o-zone possession percentage from the d-zone defending percentage we get 13.1%. Now when that number calculated drops below 12.5% team's generally give up more shot attempts than they attempt themselves.

Neutral-zone dump-ins remain an important tool in limiting neutral-zone turnovers, while placing the puck as far away from a team's defensive-zone as humanly possible. That said, any team employing the dump-and-chase model as their go-to play for producing offense will likely find themselves among the league's bottom-feeders in even-strength goals-for.