NLL Rule Book Changes: Influential or Non-factor?

During the off season, there were five new rules/modifications that were introduced into the NLL rule book. These additions were changing the time to advance the ball, changing stick length, reducing the size of goalie pads, five minute majors, and changing the substitution area rules. Each of those five things were implemented in order to effect a couple of aspects of the game.

I kept on thinking to myself and asking if all of these changes actually made a difference to the game play or if they were basically just to change the ink in the rule book. Well, I decided to look at three of the rules that were easily measurable in box score stats and see if there was any significant change. These three changes are the ball advancement time, reduction in goalie pads, and the five minute majors.

Before I opened up NLL.com to gather the stats and even before I set up my worksheet to hold these stats, I didn't know what to expect from my results. I've had people tell me that the rule changes have improved/hurt the game and I've also had people tell me that the season seems pretty much the same just with new faces and relocated players. Hopefully while you read this you learn a little about math and can start forming an educated opinion on the rule changes. At the very least, you'll have some numbers to throw around the dinner table or some answers during trivia night. Like with any respectable statistical analysis, I did not change my tests in order to support what I wanted to find so do not think that I have any bias in this write up.

I'll give you all a quick crash course in statistics after the jump and before I dive in so that you can understand what I'm writing.

Terms and Concepts

Before I get any further, I want to start off by saying that the rules affect more than one stat category and other elements have influence (i.e. the former Blazers players being distributed to other teams). However, I will be assigning the main stat that the rule affects to my analysis.

Box plot: Most of you probably know what a box plot is. The box plot shows the minimum, first quartile (the number that is the bottom 25%), median, third quartile (highest 25%) and the maximum. In addition to those five points, box plots also mark outliers (indicated by circles on the following plots) which are points that are outside 1.5 times the interquartile range (third quartile minus the first quartile). Outliers are removed from the calculations as to not influence the data in one way or the other.

Standard deviation: The standard deviation (StDev) is a measure of the spread of the data. If the StDev is at 0 then that means that all of the data points are the same and as the StDev increases that means that the numbers are further apart.

Mean: What most people call the average. The mean is the sum of all points divided by the number of points. A sample mean is what we'll be looking at for this season and is the mean of the points that we observe. The other mean is a population mean which is like the 2011 season because we have all of the points.

Confidence interval: A confidence interval (CI) is a buffer around the mean. The CI is dependent on an alpha level (I'll be using alpha = 0.05) which indicates how tight the interval is around the mean. An alpha level of 0.05 means that we are 95% (1 - alpha) confident that the population mean is within our interval.

Hypothesis testing: Hypothesis testing involves setting a null hypothesis (our guess, denoted by Ho) and an alternative hypothesis (what we are measuring against, denoted by Ha) and seeing if the data supports Ho or if we reject it in favor of Ha. For example, if we look at shots per game compared between the 2011 and 2012 season our Ho would be that 2012 shots - 2011 shots = 0 and our Ha would be that 2012 shots - 2011 shots > 0 meaning that shots have increased. The difference has to be significant in order to reject it. We use t-values to test this. T-values are dependent on degrees of freedom (based on the number of data points) and our alpha value.

Now that I've sufficiently bored you and lost half of the readers, let's dive into the good stuff. I'll avoid going full math nerd on everyone and leave the computations out of here.

8 Second Rule

Rule 47: 8-SECOND COUNT
In all situations, there shall be an eight (8) second count half-court applicable to both teams. When a team gains possession in the defensive half of the floor, they must get the ball across the centerline within eight (8) seconds of gaining possession.

This rule has the major effect of increasing the pace of the game which would hopefully lead to an increase in shots. Simple logic leads us from standing in front of the goal (in the attacking half) would lead to more time to take shots. Let's check if this actually happened.

In Figure 1.1 we have a comparison of the shots (SOG and SOFF) per team per season and Figure 1.2 indicates the total shots per game. As you can see, the median doesn't change much in either figure. In fact, the mean has only risen from 64.7 to 66.2 which isn't that much of a increase (2.3%).

The 95% confidence interval for shots per team are (62,67) and (62,70) for the 2011 and 2012 seasons, respectively. Our Ho is statistically supported by our findings and we are led to believe that the population mean for this season is approximately the same as the one from last season. The same findings are gathered from calculating a confidence interval for the total shots per game, (125,133) for 2011 and (127,138) for 2012.

In my opinion, if the league really wanted to increase the pace of the game, instead of dropping the time to advance the ball they could have shortened the shot clock or removed the resetting of the advancing clock after a time out. By dropping the clock from 30 seconds to 25 would stop the teams from moving the ball forward and then setting up plays. Yes, play would probably get a little more sloppy but they would get used to the changes just like they have this year.

Reduction in Goalie Pads

35.4 PANTS
The maximum width (straight line) of the thigh pad across the front of the leg is eleven inches (11).

35.5.1 SHIN PADS
Shall not exceed eight inches (8") in extreme width from the base of the knee cap when on the leg of the goaltender. The shin pad shall not exceed five inches 1/2 (5" 1/2) in width at the base of the ankle and taper in a straight line from the knee cap to ankle with conformity to leg. Shin
pad width at or above the knee cap shall not exceed eleven inches (11").

35.7 ARM & CHEST PROTECTOR
The contoured padding must not be more than three inches (3") in width outside the said goaltenders most outer body point on each side. In addition the padding may not climb higher than two inches (2") above the plane of said goaltenders shoulder. The arm padding may not measure more than seven inches across (7") at its greatest point. The contoured padding must not be more than two and 1/2 inches (2 1/2") each side of the chest and abdomen.

The purpose of removing some of the pads was to increase the mobility of the goalies while not removing all of the gear which would reduce their safety. By removing even an inch of padding, the shooters with the best aim will have more space to squeeze the ball into the corners which will increase scoring. I don't know if the NLL were expecting the goal totals to increase or not, but I can only assume that they would know that it was possible.

The jump between years is much more noticeable when it comes to scoring as opposed to shots. The mean number of goals has jumped around 1.3 goals per team (11 to 12.3) and 2.6 for the game (21.9 to 24.5). The StDev has remained pretty similar between both years so I am led to believe that goals has actually increased. If we do hypothesis testing with Ho being 2012 goals - 2011 goals = 0 and Ha being 2012 goals - 2011 goals > 0, we reject our null hypothesis in favor of our alternative for both team goals and game goals.

While mobility has increased (Pat Campbell went from the crease to the back wall to start a fight faster than I've ever seen before), the goal tallies have increased also as there is less material filling up the pipes. As long as safety is remains the same, I am all for this rule because it allows better/more acrobatic saves but goal totals also increase.

Five Minute Majors Released by Three Goals

38.1 MAJOR PENALTY -
For a "Major Penalty" in any game, the offender shall be ruled off the floor for five (5) minutes. If three extra man goals are scored on the penalized team, the penalized team may place a player on the floor to get back to even strength.

The final rule change that I'm looking at is changing the major penalty rule from two goals equals even strength to three goals before a runner comes back on. The increases the severity of penalties and discourages players from attempting to hurt the opposition. Because we are dealing with penalties, what better stat to look at then power play goals.

Much like with the goalie pads, we see an increase from 2011 to 2012 in the number of power play goals. Teams are now averaging 2.763 PPG per game when in 2011 they were at 2.075. Per game totals have jumped from 4.15 to 5.526. Because of the small sample size this year, there is an increase in StDev for both categories.

So far this season, we have seen three games with double digit PPG (Buf/Min - 11, Roc/Phi and Col/Cal - 10) out of a total of 19 games while last season's 80 games only saw a single double digit PPG game (Col/Cal). As you can see in Figure 3.2, the median increased from four up to five as it matched Q3 from last season. The 95% CI for this season is (3.9,7.1) and last season's CI is clinging to the lower tail at (3.6,4.7).

There is significant statistical data for us to reject our null hypothesis and believe that PPG has increased. Even though PPG has increased, this is more than likely one of those circumstances where we have data that supports a hypothetical increase in goals but it doesn't imply causation. With any contact sport, punishment needs to be taken seriously for any illegal hits such as a blow to the head. I believe that having the offending team play man down for three goals instead of two is good for the game in terms of increasing the severity of the punishment. The likelihood of getting back to even strength before the five minutes is even less likely now.

We have now learned that two of the three rule changes have made an effect on the game so far this season. The only that did not, eight second rule, was probably due to it being trivial in nature. Eight seconds is still plenty of time to change players and advance the ball and in the rare cases that it isn't, you can just take a timeout and it resets the clock. As I suggested up in that section, more of a change would come in dropping the shot clock down to 25 seconds. Not only will that increase the number of possessions per game, but it would also force the attacking team to advance the ball forward. With the current 30 second shot clock, teams had 20 seconds or 22 seconds to sit around the goal once they advanced the ball. If the clock was dropped to 25 seconds, teams would only have 15 or 17 seconds depending on the ball advancing rules. I know that if I only had 15-17 seconds to get a shot off, I would move the ball forward in transition a lot quicker so that we would have more time to set up a shot.

I wouldn't be surprised to see a couple of new changes in the 2013 rule book but I would like to see the league do some more research if possible before putting in the new changes.