The Myth of the Game in Hand

I catch a lot of slack for this but the Game in Hand advantage is largely a myth. I say this because, especially this time of year, a lot of people lean on it like a crutch giving them false hope their team can overcome, particularly now if it’s a small deficit down the stretch trying to make the playoffs. This is particularly prevalent in the National Hockey League that uses points rather than win percentages to determine the standings. The theory goes that the team with fewer games has the potential for extra points, such as an extra win, not currently reflected in the standings.

Of course, it is only an advantage when the game actually materializes and results in points, such as a win. Otherwise, it’s just another pointless loss. But, in the waining days of the season not all point affect the standings in the same ways, thus it’s important to take the Game in Hand with context.

Whenever I bring up the need for context I typically get the reflexive, neanderthal answer of “it’s universally accepted that a game in hand is always an advantage” and then I’m told I’m stupid for justifying my feeling it’s not a universal advantage using real world data. The argument over the application of data, particularly in the eye-test world of sports isn’t going away anytime soon and there’s definitely something emotional about wanting that hope the myth of the game in hand provides.

First thing to account for is proximity to end of season. Before the final fifth of the season the game in hand is fairly irrelevant since there’s a lot of schedule noise to be considered. The advantage is washed out pretty quickly with so many standing points still available and the ongoing effects of the unbalanced schedules where teams often trade who has the GIH.

Once the final 20 to maybe 30 games are reached the Game in Hand begins to materialize into a possible advantage to the greatest number of teams in the most situations. Teams that are down in points perceive the advantage of having more points available to them on the table while still having a long enough window of opportunity to earn them over.

However within the last 10, when there’s a fixed deadline for the end of points accumulation that teams are up against. The game in hand’s advantage is greatly diminished because the window of opportunity to earn points leaves very little margin for error. For the purpose of this discussion since this is when there seems to be the most contentious talk about game in hand, and it is within the window the current season is in, we’ll limit it only to a this timeframe.

Second is the points differential. Different differentials have different implications, which is too often lost in the conversation but should be fairly obvious:

GIH Team is greater than or equal to the points of the Up-game Team then GIH is an obviously huge advantage. A win provides a cushion, a loss does little-to-no damage.

GIH Team is down by only a point to the Up-game Team then GIH can be an advantage. A win pushes the GIH team ahead, an OTL ties and a loss might not hurt as much as a single standing point can come from just an OT appearance.

GIH Team is down by two points to the Up-game Team then GIH can also be an advantage but also leans disadvantage. A win produces a tie, that’s good. However, an OTL closes the gap a little but still hurts as you’ll see in the next examples about the 1-point but even games conundrum. A loss hurts.

GIH is down by three to the Up-Game Team then GIH probably provides very little advantage. A win does close the gap, but still leaves the team trailing by one. While I conceded one behind is better than three the problem mathematically still arises from the standing points-per-game and how the magic number is ultimately affected over a ten or fewer game run. This would be more complicated with an OTL being 2 back. And a loss becomes a detriment.

GIH is down by four or more to the Up-Game team and the advantage is actually almost nothing. A win again closes the gap but not enough to make a much of a realistic difference in a short run and the OTL does even less. A loss pretty much begins to dash all hopes.

It’s that weird three points back with under 10 games to play that seems to tick people off the most though. So, let’s use some real world examples to demonstrate how winning the game in hand still leaves the GIH Team at a substancial disadvantage by looking at the situation between the Ottawa Senators and the Boston Bruins as they fight for the final Wild Card spot in the East this season.

Ottawa is the GIH team and are three standing points behind the Bruins. The Sens have 7 games remaining, or a maximum 14 points on the table. If they earned at a 100% rate this brings them to 100 points.
The Bruins have six games remaining, or a maximum of 12 points on the table. They only need need 11 to reach the same 100 points, a perfect run nets the win.
If the tie occurred it’s important to note the tie-breakers since a tie alone isn’t good enough for Ottawa to achieve success. The first tie-breaker is ROW, in which the Bruins hold a 3 game advantage. During the final stretch, if Boston earns 0 ROW then of those 7 wins Ottawa needs, they must earn 4 ROW to win outright or 3 ROW to force the second tiebreaker of Head-to-Head season record where Ottawa has a 3-1 advantage. For every additional ROW Boston picks up the minimum required for Ottawa increases.

Probably is neither team is going to sweep their final games. The longest points collection streak so far this year for year team doesn’t collect the points needed and we’ll assume over a 70+ game sample their body of work reflects their general abilities to collect points.

However, the opposite is probably not very likely either, which is Boston earning zero points over the final games which only requires Ottawa earn four standing points to jump, something like a minimum of 2-4-0, 1-2-3 or 0-2-4. While Boston has losing streaks that long, they’ve collected some OTL points during it and again with a 70+ game sample size there’s a bit of an assumption the scenario doesn’t happen.

So, watch how each point Boston earns though removes the positive effect of an outright win on the Game in Hand keeping in mind the discussion about differentials earlier.

Boston earns only a single point, Ottawa needs at least five standing points. Winning the game in hand picks up ground and leaves only three remaining standing points to earn in the run which is the same distance behind as they were before. Losing still leaves six games to pick up all five points. However, about a point per game pace is manageable.

Boston earns two points, Ottawa needs six. The game in hand win allows Ottawa to keep pace which appears to be a temporary boost in the standings, but they still need four standing points over their final six games, or a point per game pace. Again, probably manageable, but the affects of the GIH win are diminished because they’re still playing from behind.

Boston earns three points, Ottawa needs seven. The game in hand win has Ottawa again leaves Ottawa temporarily in a better position but because they were playing from behind to begin with the effect begins to diminish more. Now Ottawa is playing at slightly greater than a point per game pace over the final six.

Boston earns four points, something like a 2-4-0 or 1-3-2, now Ottawa needs at least 8, or a 4-3-0, 3-2-2, 2-1-4 or 1-0-6 type record to jump, playing at a 1.33 points per game pace. That’s the equivalent of playing at a 109 point pace over an 82 game season. While that type of steak isn’t unheard of, especially for desperately hot teams this time of year, it’s also not probably, which is why Ottawa is on the “death watch” lists.

Each game Boston picks up points above four, especially in the form of ROW, makes it increasingly more difficult for the Sens to catch up even having won the GIH because the GIH win still left Ottawa playing from behind. When taking the extra game into account in these scenarios: Winning the game in hand outright in regulation is the only way in which Ottawa can just keep pace. Losing it, as you would expect nearly kills Ottawa’s chances without a lot of help from Boston’s opponents

You can do the same analysis of the Washington Capitals three point deficit to the Islanders for the third spot in the Metro where the Caps have a game in hand.

Third is the probability of winning based on the schedule ahead.

First, let’s cover schedule density. Generally, there is plenty of evidence that in 10-game increments teams playing more games in fewer days tend to collect fewer standing points than those teams who have more days of rest during the given 10-game segment. In the final stretch there’s a fixed end date of the last day of the season. The team with the game in hand would appear to be at a disadvantage within the ten game window because they have fewer days to complete their ten games than the team that has the Up-game and only 9 left to play. When taken in the context of a points-per-game those additional standing points could be harder to come by because of the grind of playing more games in less time. Density though is more than just a function of number of games in a given number of days but the layout of those games themselves. Evenly spaced games appear easier to pick up points in than combinations of long layoffs with back-to-backs, meaning the same number of games in the same number of days doesn’t always have the same effect on the ability to pick up points.

Second, is not all games are equal so, there’s Strength of Schedule. This uses a head-to-head comparison of each teams remaining games to determine the assumed difficulty of the games and the likelihood of winning. Different places calculate opponent difficulty in SoS in different ways, but an easy to understand variation of calculating the opponent’s difficultly is standing points per game/position in the standings, goal differential, recent trend of the opponent, and previous record against the opponent (or as a proxy, record against similar teams in the same conference).

Now, there are some teams who despite their place in the standings always play other teams very hard. Think Philadelphia who is out of the playoff hunt facing a Metro opponent like the Rangers or Pittsburgh. While Philly would appear on paper to be a weaker opponent it is not an easy win for the Rags or Birds which is why including the last component is popular.

Third is recent trend of the team itself. There is some evidence to support the concept that the cumulative outcome of a previous segment of games can help predictive the cumulative outcome of a forthcoming series of games. Therefore, on a rolling ten game basis it can be possible based on the standing points earned to predict the forthcoming ten game earning rate based on the previous ten game earn rate. This is, of course, assuming all else about the 20 games in question is the same, barring injuries, suspensions or other substantial changes to the franchise dynamic. For those inclined to use deeper analytics which are different indicators of future success than just the accumulation of standing points, Fenwick plots are quite common in deducing the direction in which a team is trending based on their ability to possess the puck. Thus, hard luck losses are wiped out of the equation when attempting to predict the likelihood of future success.

For all three of these, I could go back and apply them to the Ottawa – Boston scenario. However I’m not a Boston or Ottawa fan. What’s more interesting to me is the situation between the Caps and Islanders. The two teams are essentially in the same position as Ottawa and Boston are. Before tonight’s game they were

Advertisements

About thedoormouse

I am I. That’s all that i am. my little mousehole in cyberspace of fiction, recipes, sacrasm, op-ed on music, sports, and other notations both grand and tiny: https://thedmouse.wordpress.com/about-thedmouse/
This entry was posted in Opinion, sports commentary. Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s