## Inspiration

Many programs predict what existing players next season statistics will be, but can you predict what a players next season statistics will be when they aren't already in the database? How do you predict how well a player will do if you're hiring them?

With 'Pitch Your Player', you can input custom statistics to predict how well the player will do next season. If you're a hiring manager for a team, this would be invaluable.

## What it does

Choose whether you want to look at batting predictions or pitching predictions. Then choose the input parameter: age of the player, years the player has been playing, and 3 other overall parameters based on whether you are looking at batting or pitching predictions. Batter looks at '_ on base percentage + slugging ', ' on base percentage ', and ' batting average '. Pitcher looks at ' earned run average ', ' opposing batting average ', and ' strikeouts per 9 innings _'.

Using these values as inputs, our model then predicts what the players' corresponding statics will be for the next season.

## How we built it

We used python to create our program and flask to make it into an API.

## Challenges we ran into

Initially we wanted to create a neural network that would be taught by online professional baseball player statistics and create an accurate model that could predict what a person's next season statistics were, based off of previous years.

## Accomplishments that we're proud of

The model that we used as a prototype for what our future model would be was very practical but also elegant. Most players statistics are relatively close to what they have been in the past, but two factors seem to cause this difference to increase; age and how many years the have been playing. Therefore, the greater the age is minus a portion of the years played, determined the random difference of the next season's statistics. The elegant part was the we ensured that our results would always be the same if the user input the same statistics. We did this by using the sum of all the inputs from the user as the seed for our random number generator.

## What we learned

We learned to create a website, code in javascript, connect a Python api, and work together as a team.

## What's next for Pitch Your Player

The next steps would be to develop the neural network and use that to make the predictions. This would make our model more robust and accurate, and give more information to the customer. A customer could enter data for every year their player has played. The neural network would then make the predictions made off of that live information.