As I’m sure you know, last week I tested out NHL 15’s real world prediction powers by having it simulate 16 games (at least 1 per team in the NHL) between Friday 24th and Sunday 26th. I stuck to the same rule set I used for the FIFA 15 predictions with stock rosters and default skill levels.
However I did decide to mix up the way I’d score the predictions. Instead of offering a point to NHL for each correct scorer, result and method of scoring I’d instead base the results on score alone. 2 points will be awarded to NHL for a correct score line and 1 point for getting half the score line correct. The reverse will be true for reality, if NHL predicted the complete score incorrectly then reality will score 2 points, a single team score predicted incorrectly will result in reality gaining a single point. Pretty simple right? Out of 16 games predicted NHL could walk away with a maximum of 32 points, so let’s crack on with the scoring.
|
Prediction |
Reality |
Points Awarded |
|||||||||
|
NHL |
Reality |
||||||||||
|
FRI 24 OCT 2014 |
|||||||||||
|
Vancouver |
0 |
5 |
Colorado |
|
Vancouver |
3 |
7 |
Colorado |
|
0 |
2 |
|
Carolina |
8 |
1 |
Edmonton |
|
Carolina |
3 |
6 |
Edmonton |
|
0 |
2 |
|
Columbus |
0 |
1 |
Anaheim |
OT |
Columbus |
1 |
4 |
Anaheim |
|
0 |
2 |
|
SAT 25 OCT 2014 |
|||||||||||
|
Buffalo |
0 |
1 |
San José |
|
Buffalo |
2 |
1 |
San José |
|
1 |
1 |
|
Boston |
1 |
3 |
Toronto |
|
Boston |
4 |
1 |
Toronto |
|
0 |
2 |
|
NY Rangers |
0 |
3 |
Montreal |
|
NY Rangers |
1 |
3 |
Montreal |
|
1 |
1 |
|
New Jersey |
4 |
1 |
Ottawa |
|
New Jersey |
3 |
2 |
Ottawa |
OT |
0 |
2 |
|
Dallas |
1 |
0 |
NY Islanders |
|
Dallas |
5 |
7 |
NY Islanders |
|
0 |
2 |
|
Detroit |
2 |
3 |
Philadelphia |
OT |
Detroit |
2 |
4 |
Philadelphia |
|
1 |
1 |
|
Chicago |
0 |
2 |
St. Louis |
|
Chicago |
2 |
3 |
St. Louis |
|
0 |
2 |
|
Pittsburgh |
0 |
3 |
Nashville |
|
Pittsburgh |
3 |
0 |
Nashville |
|
0 |
2 |
|
Tampa Bay |
1 |
0 |
Minnesota |
|
Tampa Bay |
2 |
7 |
Minnesota |
|
0 |
2 |
|
Florida |
1 |
2 |
Arizona |
PEN |
Florida |
1 |
2 |
Arizona |
OT |
2 |
0 |
|
Washington |
1 |
2 |
Calgary |
|
Washington |
3 |
1 |
Calgary |
|
0 |
2 |
|
SUN 26 OCT 2014 |
|||||||||||
|
Colorado |
2 |
0 |
Winnipeg |
|
Colorado |
1 |
2 |
Winnipeg |
OT |
0 |
2 |
|
Columbus |
0 |
3 |
Los Angeles |
|
Columbus |
2 |
5 |
Los Angeles |
|
0 |
2 |
|
|
|||||||||||
|
Total Score |
5 |
27 |
|||||||||
Well that didn’t go as well as it could have gone with NHL predicting only 16% of the results correctly. FIFA managed to get 30% of the results right but had roughly 3 times the number of games to predict. The most impressive result for me is the Florida/Arizona game in which the prediction and reality almost lined up perfectly. Whilst it wasn’t won through penalties it did go to over time so was still settled outside of regulation.
So can we take anything away from this? Well yes, first unless you’re feeling really brave and can afford many losses then don’t place real world bets based solely on a video games predictions. Second is that whilst games seem to be advancing these advancements don’t necessarily make them better predictors of real events. Although granted there are many more factors in the real world than the virtual one. Finally these games will never replace going out and playing or watching the sport for yourself. Whilst they are cheaper in the long run, the atmosphere at a sporting venue or the feeling of taking a shot on net cannot be replicated by a game, yet.
As always feel free to join in the conversation using the comments below or over on Twitter where I’m @PaulMurday.
Paul Murday
Editor
GameSyncUK 