Masterlock App🏈

Masterlock Whitepaper

Summary

The Masterlock App chooses its picks based on the idea that oddsmakers are wrong somewhat consistently and predictably. The spread is the best indicator of oddsmakers valuation of a team, so instead of attempting to figure out the winner of each game, we just need to figure out if oddsmakers overvalue or undervalue certain teams in terms of their spread valuation.


Definitions:

Spread: noun The spread is an attempt of an oddsmaker to bring a matchup as close to a 50%/50% as possible. It is the number of points that the oddsmaker feels each team will win or lose by. For example, if the Browns are an underdog by -7 points, meaning that they are projected to lose by 7, then they will have a spread of +7.

Cover: verb To cover is to win against the spread. If a team scores more points than the other team plus their spread, they have covered. For example, if the Browns are an -7 point underdog, and they win by 10 points, they have covered the spread by 10 - (+7) = +3, which is a won bet.

Oddsmaker: noun Sometimes referred to here as "Vegas", the oddsmaker is the individual or organization who sets the spread and prices for a matchup based on their data. It is in the oddsmakers best interest to make bettors lose as many bets as possible.

Price: noun The price is the multiple of winnings from a bet. A spread's price is usually -110. This means that if you bet $11, your payout would be $10. It isn't exactly even so that the oddsmaker can make money more consistently.

Background

The idea of Masterlock came to be as a result of a group of friends who were avid sports bettors. The hypothesis of Masterlock was based on the idea that a team who more consistently has covered the spread would continue to do so in the future.

The second hypothesis was that a team who consistently did not cover the spread would continue to do so in future games.

This hypothesis is based on the concept that oddsmakers are never 100% correct; they use a sophisticated dataset to build a highly accurate model, and the effort to recreate such a model is too much for most non-enterprise teams.

The catch, however, is that this data leads to predictable outcomes, so we don't need that model. We need to track their model. The spread of any given matchup is as close to a 50-50 bet as an oddsmaker can produce. This can be deduced from the price of these bets being around -110 on either side (which we will consider even for this explanation).

This leads to the reasonable assumption that a team that has covered consistently is undervalued in the oddsmakers model, since the oddsmakers are consistently estimating their score to be lower than it actually is.

Implementation

In order to take advantage of this predictability, our solution is multifaceted:

The first step is to gather matchups for the given day. We use a simple API to gather and clean that data.

The second step is to gather each team's win percentage against the spread. We're going to call this their ATS Record, or ATS Win Percent. We calculate this by taking their number of games that each team has covered, divided by the total number of games they have played. We do not count a push (or a tie) as a win.

Once we have both of these sets of data, the third step is to compare, for each matchup, each team's win percentages and find the difference. We subtract the two win percentages and then take the absolute value of the difference.

Now that we have the win percent differential (which we will refer to as just "the differential"), step four is to sort them in order of highest differential and assign each a label. The best possible label is a Masterlock. This is given to differentials of >55%. Then we have Great, Good, Decent, and Cringe, in descending order of differential.

Due to the price of spreads, in order to consistently be making money, we need a win percentage of >55%.

We can take this differential into account to make picks. In the future, we will be implementing a way to see how a team's spread has changed and its trend. If a team is trending towards covering by more points per game, then we can assign them a higher score.

Solution

As bettors, we can use this data to help us make picks.

We'll never be 100% correct, because you can't predict things that happen during every game, but as long as Vegas is consistent, we should be able to win.

At the end of the day, we'll always need to come together as a community to decide what looks good and what doesn't. That's why the Locksmiths exist, so if you'd like to be a part of it, join our Discord, and let's start picking some locks.