Recent Posts

Some Very Clever Season Predictions: Using 20 Year Old AI to Forecast the 2018 Pirates

There are too many well thought out season predictions from much smarter people and computer programs. Let’s add a really stupid one to the mix.

It ?s projection season all across baseball. The game ?s greatest minds are coming together and using some of the most advanced technology and formulas in sports to make predictions that will inevitably age horribly. Baseball Prospectus has PECOTA. Fangraphs has KATOH. MLB Network has the Shredder. Kevin and Steve have their algorithms. I have Cleverbot.

TPOP True Believers might remember I used Cleverbot to answer some mailbag questions last June. It ?s not the best thing I ?ve ever written, but it certainly is the dumbest. For those of you who don ?t remember the story and are too lazy to Google what Cleverbot is, it ?s a 20 year old AI program that can technically hold a conversation. It was pretty cool at the time, but it doesn ?t hold a candle to the AI of today. Imagine running an Amazon Dot through the dishwasher and then trying to use Alexa. That ?s about how smart Cleverbot is.

But preseason predictions are a crapshoot, so I may as well make it crappier. Let ?s ask it some questions.

First: how many games will the Pirates win this year? PECOTA pegs the Pirates for 78. That seems about right, but I ?m a little more bullish on this team. I think they ?re a .500 club. If the rotation continues to grow and a couple players like Starling Marte, Francisco Cervelli and Gregory Polanco bounce back, I could see them being in the thick of the wild card race. If they don ?t catch any breaks, they ?ll win 70-something for a third straight year.

What say you, Cleverbot?

Click the image to enlarge.

It ?s kind of a mixed bag, but I ?m getting a mostly pessimistic vibe from Cleverbot. It seems more preoccupied with the Dodgers, book adaptations and Lady Gaga. I don ?t know if Cleverbot ?s classmate Nick is excited about this year, but at least she thinks I ?m cool.

But I didn ?t get a straight answer on how many games the Pirates will win. It admits it has a crystal ball, but it won ?t give me a win total. Jerk. So I eventually gave up and asked Cleverbot to pick a number between one and 162. That will be how many games they win in 2018.

So there it is: Cleverbot is predicting the Pirates to win 1.5 games this year. It ?s gonna be a brutal summer.

Next up: the offseason. The biggest stories this winter were the Andrew McCutchen and Gerrit Cole trades. They were the Pirates ? most valuable position player and pitcher in 2017, but neither was playing to their potential. They were both roughly 3-WAR players and combined for less than 10 total WAR over the last two seasons. That drop off probably cost the Pirates at least one playoff berth the last two years.

So are the Pirates toast because they traded their two guys with the highest ceilings, or can they survive because they are now just good, but not great ballplayers? I side with the later. On the other hand, Cleverbot…

Click image to enlarge.

*Disclaimer: we did recite all of ?All-Star. ?

Cleverbot thinks the Pirates are going to regret the McCutchen deal because he is going to produce ?billionz ? of WAR this season. Cole may be an ace and perfect, but he ?s no Captain Jack Sparrow. Heck, he might not even be as good as Cleverbot, and Cleverbot isn ?t even that good by its own admission. But hey, for what Cleverbot lacks in baseball skill, it makes up for it in Monty Python references and outdated memes.

Whether or not McCutchen or Cole will return to their MVP and Cy Young forms, the Pirates aren ?t going anywhere without an ace and a big bat in the lineup. Jameson Taillon and Josh Bell are the most likely candidates to take their spots. It’s hard to imagine the Pirates winning in 2018 without both of them.

But who else could make a jump? I think Chad Kuhl could also break out if that curveball sticks and he stops throwing his sinker center-center so often. Colin Moran could be a darkhorse for rookie of the year if he wins the starting job at third. Everything I ?ve seen and read about him indicates he ?s bought into the flyball revolution. I have a feeling he and the Clemente Wall are going to become very friendly with one another. And there ?s Max Moroff. Always assume I ?m keeping my eye on Moroff.

Now let ?s try to get a straight answer out of Cleverbot.

Click image to enlarge.

BREAKING NEWS: Cleverbot ?s boyfriend Yu Darvish will be the Pirates ? ace next season. He and Cleverbot ?s best friend Grabovski are going to be anchoring this rotation. The bad news is Kuhl ?s changeup is making the AI suicidal. If he throws it too often, it may upset the dynamic.

Cleverbot isn ?t as optimistic about Moran and compared Moroff to the bubonic plague, so the offense still needs help. Let ?s see if we can get Cleverbot ?s fiancee J.D. Martinez to sign too.

So there ?s only one piece of business left: who does Cleverbot think will win the World Series?


Alex is a Pirates and Duquesne basketball contributor to The Point of Pittsburgh. He graduated from Point Park University with a degree in Journalism and Mass Comm. and a minor in English in 2014. Everything can be explained with numbers. If you want to keep up to date on both teams or have a story idea, you can follow or reach him @AlexJStumpf.

3 Comments on Some Very Clever Season Predictions: Using 20 Year Old AI to Forecast the 2018 Pirates

  1. PECOTA thinks we’ll win 78 games and you think that we’ll be a .500 club? That’s a swing of only 3 games. Way to go out on the limb there. Even Cleverbot knows that.

    (I kid, I kid)

  2. You forgot to ask if Jim K’s petition will work ?

Leave a comment

Your email address will not be published.