Hockey Analytics: Project Helium – Regulated Goalie Value Changes

Hockey Analytics: Project Helium – Making Adjustments to Regulated Goalie Value (RGV) Formula

After doing a pair of studies on Winnipeg Jets goalie Ondrej Pavelec, I noticed that there were major flaws with Regulated Goalie Value (RGV) hockey analytics formula.

With that being said, I am going to make the corrections necessary in order for RGV to be a valuable formula to evaluate a goaltender. Also, just so you understand what they were changed from, I am going to have an asterisk (just one) at the bottom with the old variable and it’s value.

Click on the link to see the intro to RGV article (in case you don’t understand everything).

**For those variables with no changes, there will not be any explanation underneath.

(a)Shootout loss goals (0.9613)

A1:+(-0.002) 1-0 loss

A2:+(-0.02001) 2-1 loss

A3:+(-0.0303) 3-2 loss

A4:+(0.040201) 4-3 loss

A5:+(0.05002) 5-4 loss or more goals allowed
(Odd man rushes)

More puck prose: Stanley Cup Playoffs: Eastern Conference Round 2 Preview

(b)Clean Breakaway (0.735)

B1:+(0.03002) pressure from behind

B11:+(0.20102) if goal given up in last 10 minutes of 2^nd Pd.

B12:+(0.300201) if goal is given up in the first 10 minutes of 3^rd Pd.

B13:+(0.40102) if goal is given up in last ten minutes of 3^rd Pd. or OT

B14:+(0.010306) if goal is given up in Game 6 or 7 of a playoff series. Add C14 to any of the above.

More puck prose: Stanley Cup Playoffs: Western Conference Round 2 Preview

“c”2 on 1 (0.894)

Apr 7, 2016; Boston, MA, USA; Boston Bruins goalie Tuukka Rask (40) squirts himself with water during the first period against the Detroit Red Wings at TD Garden. Mandatory Credit: Bob DeChiara-USA TODAY Sports

C1:+(-0.020108) score on wide open net

C2:+(-0.030106) score on multiple rebounds

C3:+(0.0506) score with defender involved and goalie in position

C11:+(0.20203) if goal is given up in the last ten minutes of 2^nd Pd.

C12:+(0.306008) if goal is given up in the first 10 minutes of 3^rd Pd.

C13:+(0.50103) if goal is given up in last ten minutes of 3^rd Pd. or OT

C14:+(0.0203061) if goal is given up in Game 6 or 7 of a playoff series. Add D14 to any of the above.

More puck prose: NHL Mock Draft 4.0: Rounds 1-7

(d)3 on 2 (0.9293)

D1:+(-0.00902) score with defender screening goalie

D2:+(-0.000604) score with defender/offensive player tipping puck (you can add E1 and E2 together)

D3:+(0.08408) score from beyond the faceoff circle with defender screen/puck being tipped

D11:+(0.301021) if goal is given up in last 10 minutes of 2^nd Pd.

D12:+(0.40203) if goal is given up in the first 10 minutes of 3^rd Pd.

D13:+(0.501901) if goal is given up in the last 10 minutes of 3^rd Pd. or OT

D14:+(0.03105) if goal is given up in Game 6 or 7 of a playoff series.

More puck prose: Should NHL Be Accepting of Alex Radulov Return?

(e)Neutral Zone goals (1.6171)

Mar 31, 2016; Saint Paul, MN, USA; Ottawa Senators goalie Craig Anderson (41) waits for the faceoff in the second period against the Minnesota Wild at Xcel Energy Center. the Ottawa Senators beat the Minnesota Wild 3-2. Mandatory Credit: Brad Rempel-USA TODAY Sports

E1:+(0.0043) if shot between the blue line and the red line

E2:+(0.0519) if shot before the red line

E3:+(-0.029093) if shot within 4 feet from the blue line (closest to the goalie)

E11:+(0.30406) if goal is given up in last 10 minutes of 2^nd Pd.

E12:+(0.51006) if goal is given up in first 10 minutes of 3^rd Pd.

E13:+(0.530301) if goal is given up in the last 10 minutes of 3^rd Pd. or OT

E14:+(0.05031) if goal is given up in Game 6 or 7 of a playoff series.

*(e:+1.369, E2:+0.0409, E3:N/A)

More from Editorials

The main variable needed an increase mostly due to the fact that it is too detrimental to the team to give up a goal this bad. Also, a goal given up before the red line (closer to the offensive zone of the goalie’s team) is completely unacceptable.

There is also a new variable (E3), due to the fact that I need to have a situation where value is subtracted from the first batch of variables. This will be the most heavily punished of all the variables.

More puck prose: NHL Weekly Roundup: Latest News and Rumors

(f)either face-off circle goals (0.992)

Feb 25, 2016; Toronto, Ontario, CAN; Carolina Hurricanes goaltender Cam Ward (30) during a break in the action against the Toronto Maple Leafs during the second period at the Air Canada Centre. Mandatory Credit: John E. Sokolowski-USA TODAY Sports

F1:+(-0.07041) score with defender screening goalie

F2:+(-0.00503) score with defender/offensive player tipping puck (you can add G1 and G2 together)

F3:+(+0.16061) score just above the face-off circle (in between blue line and circle)

F11:+(0.21305) if goal is given up in last 10 minutes of 2^nd Pd.

F12:+(0.22016) if goal is given up in first ten minutes of 3^rd Pd.

F13:+(0.311031) if goal is giving up in last 10 minutes of 3^rd Pd. or OT

F14:+(0.03021) if goal is given up in Game 6 or 7 of a playoff series.

*(F3:N/A)

In the case of variable “F”, I thought it would make sense to add an extra variable (F3). I made the addition because goals between the face-off circle and the blue line shouldn’t be rated the same way others are, and it doesn’t make sense to rate them as goals scored close to the net. So basically, if the goal is further than the face-off circle but not at the blue line (and also not in the high slot – could be close), add variable “F3”.

More puck prose: NHL 17 EA Cover Vote is Underway

(g)High slot goals (0.9326) – if slap shot (0.968) – if wrist shot

G1:+(-0.06032) score with defender screening goalie

G2:+(-0.00401) score with defender/offensive player tipping puck (you can add G1 and G2 together)

G11:+(0.21204) if goal is given up in last 10 minutes of 2^nd Pd.

G12:+(0.220106) if goal is given up in first ten minutes of 3^rd Pd.

G13:+(0.30102) if goal is given up in last 10 minutes of 3^rd Pd. or OT

G14:+(0.0302101) if goal is given up in Game 6 or 7 of a playoff series

More puck prose: Project Helium: Marian Hossa Regulated Plus/Minus

(h)Low slot goals (0.9215) – if slap shot (0.946) – if wrist shot

H1:+(-0.05031) score with defender screening goalie

H2:+(-0.0030011) score with defender/offensive player tipping puck (you can add H1 and H2 together)

H11:+(0.20121) if goal is given up in last ten minutes of 2^nd Pd.

H12:+(0.2023) if goal is given up in first ten minutes of 3^rd Pd.

H13:+(0.30202) if goal is given up in last ten minutes of 3^rd Pd. or OT

H14:+(0.0401003) if goal is given up in Game 6 or 7 of a playoff series

More puck prose: Project Helium: Patrice Bergeron Regulated Plus/Minus

(i)Either point goals (1.3026)

Apr 4, 2016; Vancouver, British Columbia, CAN; Los Angeles Kings goaltender Jhonas Enroth (1) and Vancouver Canucks goaltender Jacob Markstrom (25) warm up at Rogers Arena. The Vancouver Canucks won 3-2. Mandatory Credit: Anne-Marie Sorvin-USA TODAY Sports

I1:+(-0.08011) score with defender screening goalie

I2:+(-0.0060101) score with defender/offensive player tipping puck (you can add I1 and I2 together)

I11:+(0.40134) if goal is given up in last 10 minutes of 2^nd Pd.

I12:+(0.41026) if goal is given up in first 10 minutes of 3^rd Pd.

I13:+(0.501802) if goal is given up in last 10 minutes of 3^rd Pd or OT

I14:+(0.050202) if goal is given up in Game 6 or 7 of a playoff series

*(i:0.9103)

More puck prose: Project Helium: Jonathan Toews Regulated Plus/Minus

Other blue line goals (1.4062)

J1:+(-0.061101) score with defender screening goalie

J2:+(-0.00501) score with defender/offensive player tipping puck (you can add J1 and J2 together)

J11:+(0.50123) if goal is given up last 10 minutes of 2^nd Pd.

J12:+(0.510204) if goal is given up in first 10 minutes of 3^rd Pd.

J13:+(0.60308) if goal is given up in last 10 minutes of 3^rd Pd. or OT

J14:+(0.06012) if goal is given up in Game 6 or 7 of a playoff series

(j:0.9345)

Both of the blue line goals needed the main variables increased in a big way. The reason for that is because this is the easiest way to even out my hockey analytics formula. Assuming the fact that goals are scored from the blue line, it makes perfect sense to level out things by scoring these types of goals above one (value wise).

The main reason that these two variables will be more important than “F” is due to the fact that it is rare goals are scored from beyond outside the offensive zone. I am hoping this change helps create a level playing field in terms of the dynamic of the formula as a whole.

More puck prose: Project Helium: Duncan Keith vs Shea Weber Carbon Block

(k)In front of the net goals (0.8139) if take pass and score (no rebound)- any other goal in front counts as well

K1:+(-0.0301) score on first rebound

K2:+(-0.002002) score on multiple rebounds

K11:+(0.1001) if goal is given up in last 10 minutes of 2^nd Pd.

K12:+(0.185) if goal is given up in first 10 minutes of 3^rd Pd.

K13:+(0.2013) if goal is given up in last ten minutes of 3^rd Pd. or OT

K14:+(0.00204) if goal is given up in Game 6 or 7 of a playoff series.

**(all variables underneath take precedence over those in the same circumstance for variables B-K)

More puck prose: Project Helium: Taylor Hall Regulated Plus/Minus

Power Play Goals

**(l)5 on 3 (-0.030102)

L1:+(0.050602) if goal is given up in the last 10 minutes of the 2^nd Pd.

L2:+(0.049031) if goal is given up in the first 10 minutes of the 3^rd Pd.

L3:+(0.048301) if goal is given up in the last 10 minutes of the 3^rd pd. or OT

L4:+(0.0002) if goal is given up in Game 6 or 7 of a playoff series

*(l:0.03014)

More puck prose: Project Helium: Brent Seabrook Regulated Plus/Minus

**(m)4 on 3 (-0.02021)

Oct 30, 2015; Calgary, Alberta, CAN; Montreal Canadiens goalie Carey Price (31) warms up prior to the game against the Calgary Flames at Scotiabank Saddledome. Mandatory Credit: Sergei Belski-USA TODAY Sports

M1:+(0.05081) if goal is given up in the last 10 minutes of the 2^nd Pd.

M2:+(0.052611) if goal is given up in the first 10 minutes of the 3^rd Pd.

M3:+(0.05906) if goal is given up in the last 10 minutes of the 3^rd Pd. or OT

M4:+(0.000202) if goal is given up in Game 6 or 7 of a playoff series

*(m:0.020102)

More puck prose: Project Helium: Morgan Rielly Regulated Plus/Minus

**(n)5 on 4(-0.0200203)

N1:+(0.0608303) if goal is given up in the last 10 minutes of the 2^nd Pd.

N2:+(0.060903) if goal is given up in the first 10 minutes of the 3^rd Pd.

N3:+(0.06093) if goal is given up in the last 10 minutes of the 3^rd Pd. or OT

N4:+(0.0002) if goal is given up in Game 6 or 7 of a playoff series.

*(n:0.020021)

More puck prose: Project Helium: Regulated Plus/Minus Review

Even strength goals (non 5-5)

**(o)(4 on 4) (-0.010402)

O1:+(0.0608306) if goal is given up in the last 10 minutes of the 2^nd Pd.

O2:+(0.0609103) if goal is given up in the first 10 minutes of the 3^rd Pd.

O3:+(0.061836) if goal is given up in the last 10 minutes of the 3^rd Pd. or OT

O4:+(0.000101) if goal is given up in Game 6 or 7 of a playoff series.

*(o:0.01013)

More puck prose: Project Helium: Carbon Block Review

**(P) overtime situation in the regular season (3 on 3) (-0.0103061)

*(p:0.0201)

More from Puck Prose

Variable “P” doesn’t need any variables to influence it, that being due to the fact that time is the main purpose of it. Also, I needed to change one main part of variables L-P. Before, they were all positive numbers, however they are now negative. That is due to the fact that giving up a goal on a power play should subtract value from a goalies Regulated Goalie Value. Considering that when a goalie gives up a difficult goal (according to the formula), it subtracts the value accordingly.

Moving on from variable “P”, I think the changes I made the RGV have altered the formula enough to test it a few times. I want to say one thing, that being that when a goal is scored, I don’t count it the same way the NHL does. I score it by where it is shot from. As you can see in some of the variables above, a tipped puck changes the outcome of the formula. Therefore it is impossible to value a goal by who scores it but rather where the puck was shot from.

Next: Project Helium: Carbon Block - Erik Karlsson vs Drew Doughty

All in all, I think these changes could make a fairly significant difference. A lot of it is going to revolve around what game I choose to study, as I am going game-by-game at the moment. Had I not evened out the formula somewhat, it would be impossible for a goaltender to ever have a RGV above one.

Now, with these changes it is now possible depending on where the puck is shot from. Before, it wasn’t balanced so what happened is the goalie would always get a very good RGV. Now, it is possible for him to score poorly. This is still a work in progress, so stick around. More to come.

(There is no change of the rounding to the nearest ten-thousandth part of RGV)