This previous Friday morning the gentlemen behind WyoPredicts who you can find across Social Media on Instagram, Facebook, or Twitter, joined us here at SVI for the weekly First Bank of Wyoming Sports Friday to discuss their prediction model and how the math tells them who to pick in each game of the week.
We got to learn from the source itself how the statistics become more solidified as the season goes on, generally speaking the early season predictions are harder to make using the statistical model than as the year goes on which makes sense, but they do still believe that even early on, the rankings across the stat and their model tend to match up fairly nicely and they believe that the model is fairly accurate right from the start of any given season.
“Every Wednesday the state releases the stats from the previous week and we turn those over to [our statistician] who picks the most significant stats that are proven to help predict and just comparing those and then gives us a percent of the likeliness of winning”
They have now gotten to a point where they have figured out which stats factor in what percentage to a teams chances of winning any given game that they simply plug the numbers into an excel spreadsheet with a set formula that spits out said percentage for each team to win and they adhere to that formula’s pick on their social media forums. The gentlemen do concede though that occasionally in previous seasons it is not unheard of for a game to absolutely shock their formula as was said during the interview, “I know last season there were a couple of games that were like holy crap what happened there?” 1 prime example that caused some confusion and potentially some changes to the process was Farson-Eden a few years back who was picked against by the model in every game but made it all the way to the state championship game that season.
“One thing that we have to figure out is stuff like who might be out in a given week or who might be on vacation those are things that aren’t reflected in the model” stated the WyoPredicts representatives during the interview, furthering the explanation as to what outside variables may still effect their projections for a given week. The model has been nearly spot on for most of it’s existence but there is nothing that can be done about the variables of sport that naturally occur such as the reasons stated above or even just emotion or momentum that team may feel in any given week that can’t be quantified.
For games monitored on Friday by SVI, the WyoPredicts model was correct for 83% of the games in week 1, which is really solid however we can also expect that number to go up with each week further we go into the season which is the most impressive thing.
When asked how they tend to respond to student athlete engagement with their site regarding predictions WyoPredicts had this to say; “We don’t really take it personally, honestly. like we just see it as something that helps the student athletes feed their energy”. WyoPrecicts is fully aware of the effect media outlets like themselves that make these types of predictions can have on athletes and coaches who use it as “bulletin board material” throughout a week but state that they actually enjoy that aspect of what their impact is on high school sports in Wyoming.
Their week 2 predictions should be made available on all social media later this week, give them a click and a follow and engage with them the next they pick against YOUR team right before your group pulls the big upset.