How Bad Is This MLB Perfect Game Market, Really? | Presented by Kalshi

Circles Off

2026-03-04

How to Price a Perfect Game in MLB (Step by Step)

If you saw that DraftKings posted odds on whether there will be a perfect game in the MLB regular season this year, you probably saw the comments explode. People immediately point out that only 24 have ever happened, so the offered price must be terrible logic. But thinking that 1910s baseball stats should dictate today's pricing is making a huge, fundamental mistake. I want to show you a process, a framework, to price any weird rare event market based on modern conditions, not just outdated gut feelings. This method helps you move past lazy historical arguments and build a legitimate baseline.

This isn't about finding the one perfect answer, because in sports prediction, there rarely is one. Instead, we are building a structured way to test assumptions. We can see exactly what conditions need to be true for the sportsbook's line to be correct, which is far more useful. If you’re tired of seeing odds offered on events like a season long perfect game and having no idea how to analyze them intelligently, stick with me. We're going to break down this probability using actionable data points.

Here's What We'll Cover

Why using total historical perfect games is flawed modeling

How to translate initial market odds into a starting probability

The two essential components for pricing rare season long events

Calculating your baseline chance of a perfect game happening this year

Testing bullish assumptions needed to justify aggressive odds

Translating the Market Line Into Odds

Step one, no matter the market, is translating the offered price into a probability you can actually work with. When DraftKings offered Yes at plus 225 for a perfect game in MLB this season, what did that actually mean? In betting terms, plus 225 means you profit 225 dollars for every 100 dollars risked. Simple math tells us this implies a probability of about 30.8%. That is nearly one in three chances across the entire regular season. That's our starting line in the sand.

Now, it is critically important to understand the scope. We are not asking if a perfect game will happen in any single random game. We are asking if at least one occurs across the entire 162 game schedule for every team, over roughly 2,430 total games. So, that 30.8% needs to translate into a season long probability. This is where most people get lost, but our framework simplifies it.

The Two-Part Framework for Pricing Rare Events

When you look at events that might only happen once or twice in a decade, like a season long perfect game, you can usually divide the problem into two distinct pieces. This makes the math far less intimidating. For any rare event like this, we calculate:

Chances: How many real opportunities does the sport give us during the season for this event to be possible?
Conversion Rate: Given one of those real chances occurs, what is the historical percentage that the event actually finishes?

Your season probability then equals Chances multiplied by Conversion Rate. That gives you Lambda, or the expected number of occurrences per season. Think about it this way: If a team gets 10 chances, and they convert 1 percent of the time, you expect 0.1 occurrences per season. That's it. Chances times conversion equals Lambda.

Defining a True Chance: Perfect Through Six

When we talk about chances, we can't just use completed games pitched by a starter, because a perfect game is a team effort involving relief pitchers too. We need a clean, practical definition for a game where a perfect game is genuinely in play. I argue that this definition is when a team is perfect through six innings.

Why the sixth inning? Because by then, the event is on the broadcast, alerts are going out, and everyone recognizes it as a genuine bid. If perfection ends in the first inning, it's a fun stat, but it's not the same tension. Using perfect through six innings lets us ground our calculation in recent data where managers use bullpens differently than they did 80 years ago.

Based on data from the last decade or so, here is what the league generally produces:

Perfect Through Six Pitches: On average, we see about 5.8 of these bids per season.

Perfect Through Seven: This number drops significantly, landing around 2.0 times per year.

Perfect Through Eight: We usually see only about 0.9 bids that reach this stage.

So, for our first input, the Chances variable, we are anchoring with 5.8 serious bids per season.

Calculating the Conversion Rate

Next, we tackle the conversion rate. Given a team is perfect through six innings, what percentage of the time do they actually maintain that perfection for the full nine innings? This number, historically, is shockingly low. It hovers right around 1.7%.

This makes sense, right? One walk, one fielding error, one baffling pitch and the perfect game is over, even if the pitching has been historically dominant up to that point. The barrier to success once you get that far is extremely high.

Now, we combine them for our baseline expectation, the Lambda value. We multiply our chances (5.8) by our conversion rate (0.017):

5.8 multiplied by 0.017 equals approximately 0.1. That means statistically, based on recent history, we expect about 0.1 perfect game in MLB events per season.

To turn that expected count (Lambda) into a season probability, we use a standard shortcut for rare events: Probability greater than 1 equals 1 minus e to the negative Lambda. Plugging in 0.1, we land at roughly 9.5%. This is your honest, data-driven baseline for this market, miles away from the implied 30.8%.

Testing Assumptions to Justify Higher Odds

The core of sharp prediction work is testing what assumptions you, or the oddsmaker, need to believe to justify a different price. If DraftKings thinks the chance is 30.8% (their implied 225 price), their required Lambda must be much higher than our baseline of 0.1. In fact, the math shows they need a Lambda around 0.37.

How do we get from 0.1 to 0.37? It requires significant belief that either the number of chances is higher, or the conversion rate is drastically better.

Let's explore some scenarios that get you closer to that 30.8% expectation:

Scenario A: More Chances, Slightly Better Conversion. Maybe this year the league sees 7 serious bids (instead of 5.8), and the conversion rate creeps up to 2.0% (instead of 1.7%). Lambda becomes 7 times 0.02, which is 0.14. That yields a season probability of only about 13%.

Scenario B: Aggressive Bullpen Argument. Maybe you think modern relievers are so good that the conversion rate jumps up to 3.5%, but you still only see the standard 5.8 chances. Lambda is 5.8 multiplied by 0.035, equaling 0.203. This puts us around 18.5% season probability. Not quite 30.8%.

Scenario C: Hitting the Bookmaker's Target. To hit that implied 30.8%, you would need something like 8 serious bids per season AND a conversion rate near 4.6%. Or, you need the standard 5.8 bids, but the conversion rate must be around 6.4%. That conversion rate is almost four times the historical average. That's a monumental swing.

When you see the argument that the oddsmakers are wrong because there have only been 24 perfect games ever, you realize they are arguing against an emotional anchor, not against the process. To believe the plus 225 line is correct, you have to be arguing that something fundamental about modern pitching or luck suggests a conversion rate or frequency of bids far beyond what the last decade shows us. That's why, process-wise, plus 225 feels aggressive.

Common Questions About Pricing Rare Baseball Events

What Does This Framework Tell Me About Historical Data?

This method doesn't ignore history, it contextualizes it. The 24 historical perfect games provide the ultimate ceiling, but they are flawed predictors because the game environment, training, and pitching usage have fundamentally changed since the Titanic era. Your measurement of 'chances' must reflect today's environment, not 1905's environment. Using the last decade of data for chances and conversion rates gives you a much more relevant baseline for pricing a perfect game in MLB today.

How Many Perfect Through Six Bids Should I Assume Per Year?

My established estimate, rooted in recent league averages, is about 5.8 bids per year that reach the sixth inning perfectly intact. If you feel the talent level this season is substantially better or worse than average, you can adjust that input up or down. This is where you inject your expert opinion into the model. But I recommend starting with 5.8 and seeing how far you need to stray to justify current odds.

The Easiest Way to Start Today for a Perfect Game in MLB Market

If you're looking to apply this framework to a new rare event, always start by defining your 'Chance' event clearly. Is it a hat trick, a shutout, or a specific player hitting a threshold? Once you define the halfway point—the 6-inning equivalent—find the recent frequency of that halfway point occurring. That’s your Chance variable. Then, find how often that halfway point successfully concluded in the past. That's your Conversion Rate. Multiply them. That's the process.

Why Is My Baseline Probability So Low Compared to the Book?

Your baseline probability reflects what should happen if the league maintains its recent historical performance parameters. Bookmakers price higher because they are pricing in variance, or they believe one of the inputs (chances or conversion) is underappreciated by the general public. To justify their price, you need to identify a significant positive shift in the game itself that pushes your Lambda value far above the historical mean of 0.1.

Your Next Steps

We successfully built a structured, repeatable process that moves beyond simple counts of historical anomalies. You now have a method to evaluate any long shot bet by calculating Chances multiplied by Conversion Rate to find your baseline Lambda. Remember, the plus 225 price required a massive leap of faith in improved conversion or an unprecedented number of perfect bids this year, which the underlying 5.8 chances and 1.7% conversion don't support.

Don't just walk away with the math. I want you to engage with the assumptions. Go look at the data—how many perfect through sixes did we really see last year and the year before? Then, tell me in the comments what your personal, adjusted conversion rate is. Do you think relievers make it easier or harder to finish? If you want to see more deep dives like this, hit that subscribe button so we can price out the next wild market that pops up.

About Circle Back

To support Circles Back: Sign up for new sportsbook accounts using our custom links and offers. Click HERE.

Stay Updated: Subscribe for more Circle Back content on your favourite platforms:

YouTube | Spotify | Apple Podcasts

Follow Us on Social Media:

🔨 Sign up to Kirk's Hammer

Scale Your Winnings With Betstamp PRO

Betstamp Pro saves you time and resources by identifying edges across 100+ sportsbooks in real-time. Leverage the most efficient true line in the industry and discover why Betstamp Pro is essential for top-down bettors.

Limited number of spots available! Apply for your free 1-on-1 product demo by clicking the banner below.

Episode Transcript

[00:00] DraftKings posted a very interesting

[00:02] market this week that had people

[00:03] outraged. Will we see a perfect game in

[00:06] the MLB this regular season? Yes, is

[00:09] plus 225. And the comments are going

[00:12] exactly where you'd expect. A whole lot

[00:14] of, well, there's only been 24 perfect

[00:16] games ever. So, this number is a really

[00:18] terrible price. Here's the problem with

[00:20] that logic. If you're using perfect

[00:23] games from the early 1900s to predict

[00:27] what happens today, you're pretending

[00:29] that baseball hasn't changed at all. I

[00:32] don't know about you, but I don't think

[00:33] that games that were played when the

[00:35] Titanic was operational should have any

[00:38] bearing on how we predict baseball games

[00:41] today. Think about it. Back then,

[00:43] starters used to finish games all the

[00:45] time. Bullpens weren't modern weapons.

[00:48] pitch counts weren't managed as much as

[00:50] they are nowadays. The entire runcoring

[00:53] environment was completely different.

[00:55] So, the point isn't that history doesn't

[00:57] matter. The point is that you can't

[01:00] price today's environment using

[01:03] yesterday's conditions. What I want to

[01:06] do here is not pretend like there's one

[01:08] correct answer to pricing this. I just

[01:11] want to show you a process, a framework

[01:13] that you can use for any weird rare

[01:16] event type of market. And quick favor

[01:18] before we get into it. If you like these

[01:20] types of videos, price the market type

[01:22] of videos, hit like, hit subscribe. If

[01:24] this one does well, we'll keep doing

[01:26] them going forward. All right, let's

[01:28] price it.

[01:30] Here we go. All right, we're back here.

[01:33] Step one is always the same. Translate

[01:36] things into odds. Plus 225 implies about

[01:40] a 30.8%

[01:43] probability. Roughly 1 in three. So

[01:46] that's our starting point. plus 225

[01:48] 30.8%.

[01:51] Simple. And just so we're clear, this is

[01:54] not does a perfect game happen in a

[01:56] random MLB game. This is a seasonl long

[01:59] question that we're answering. So over

[02:01] the entire regular season, do we get at

[02:04] least one perfect game anywhere? So

[02:07] we're going to need a season level

[02:09] probability. Now, here's the process.

[02:13] This could be different for different

[02:15] people, but for me, when you price rare

[02:16] events across a season, you can usually

[02:19] break it down into two parts. I'm going

[02:20] to break this one down into those two

[02:23] parts. Part one, how many real chances

[02:25] does the sport even give you? And part

[02:27] two, given a real chance, how often does

[02:31] the outcome actually happen? That's it.

[02:35] Chances times conversion rate. Simple.

[02:38] Just remember that. So, our our

[02:40] percentage for the season equals chances

[02:43] times conversion with the worst

[02:46] penmanship imaginable. All right. Now,

[02:49] for the market, I I don't want to define

[02:51] chances as complete games because that's

[02:53] a starters only lens. A perfect game is

[02:56] a team event. It's 27 up, 27 down. You

[03:00] have your starter, bullpen, whatever.

[03:02] So, the clean way to define a chance is

[03:05] a game where a perfect game is genuinely

[03:07] in play. And here's a simple practical

[03:10] definition that I'm going to use. A

[03:12] chance is a game where a team is perfect

[03:15] through six innings. Because if you're

[03:17] perfect through six, it it feels real.

[03:20] It's on the broadcast. You might get a

[03:22] push alert. It's on everyone's head. And

[03:24] one tiny thing ends it. So chances

[03:30] equals perfect

[03:32] through six. All right. Now, this is

[03:36] where we can make the script more

[03:37] grounded. Over the last decade, there's

[03:40] data on how often teams are perfect

[03:42] through six and what happens after that.

[03:45] On average, across the season, the

[03:47] league produces about five to six

[03:50] perfect through six types of bids. All

[03:53] right, it it's roughly 5.8 per season

[03:57] that we're going to use as

[04:00] our rough estimate here. Then it drops

[04:02] off a cliff. Perfect through seven. That

[04:05] happens around two times per season.

[04:07] Perfect through eight is around one time

[04:10] per season. So through seven

[04:15] equals 2.0

[04:16] through 8

[04:19] equals roughly 0.9. All right. So now

[04:22] we've got our first input chances.

[04:27] That's the first half of our equation.

[04:29] The second conversion.

[04:31] Given that a team is perfect through

[04:33] six, what percentage of the time do they

[04:37] actually finish it off? And this number

[04:40] shockingly very small. Historically,

[04:43] it's about 1.7%.

[04:46] In other words, if you hear perfect game

[04:49] mentioned in the sixth inning,

[04:51] statistically

[04:52] it usually doesn't end well. All right.

[04:56] So, 1.7%

[04:58] for our conversion here roughly. And

[05:01] again, we'll get into this a little bit

[05:03] deeper as we go on. Now, we just combine

[05:06] those two inputs. Expected perfect games

[05:09] per season is what statisticians would

[05:12] call lambda. Lambda is just chances

[05:16] times conversion. So, if chances are

[05:18] about 5.8 8 and conversion is about

[05:21] 1.7%.

[05:23] Then we have lambda and that's equal to

[05:27] 5.8

[05:29] time 0.017

[05:32] and that is approximately 0.1.

[05:36] Okay. Now we translate lambda into a

[05:39] seasonal probability. So lambda is one

[05:41] of our variables here. Here's the

[05:42] standard shortcut for rare events. Okay.

[05:45] You don't have to remember this. I'll

[05:46] write it down. But the probability of at

[05:48] least one equals 1 - e to the lambda.

[05:53] Okay, again sounds confusing but

[05:57] probability of greater than 1 is equal

[06:00] to 1us e to the lambda. Simple equation.

[06:07] Okay. So if lambda is about 0.1 1 minus

[06:11] e to the 0.1 is about 9 to 10%. All

[06:16] right. So, if we're using this and we're

[06:17] plugging it in, we get approximately 9

[06:20] and a half%. All right. Now, pause right

[06:24] here because this is the whole point of

[06:25] doing this process first. If you anchor

[06:28] your chances and your conversion in

[06:31] actual data, your baseline here comes

[06:34] out closer to 10%. Not the 30.8 that we

[06:38] got earlier. So, how does anyone get to

[06:41] a bigger number? And this is where you

[06:43] do the honest part. You test what

[06:45] assumptions would need to be true to

[06:48] change that number because the league

[06:50] isn't static. Maybe you believe that

[06:52] modern pitching quality makes perfect

[06:54] through six happen way more often than I

[06:57] projected. Or maybe you believe that

[07:00] bullpens nowadays, relievers throwing

[07:02] 104 miles an hour make finishing

[07:05] slightly easier once a perfect game bid

[07:08] is alive. So we can do some scenario

[07:10] checks. Scenario one, maybe we get seven

[07:14] serious bids instead of our 5.8.

[07:18] Maybe conversion is higher. It's not

[07:21] 1.7%, maybe we bump it to 2%. Then we do

[07:24] lambda 7 times. 02, it's 0.14. That's

[07:29] still only around, you know, 13 to 14%

[07:33] for the entire season, right? You can

[07:35] have an even more bullish world than

[07:37] that. Let's say you have eight serious

[07:40] bits instead of roughly six this year

[07:44] and conversion is even higher. It's two

[07:46] and a half percent. 8 times 0.025 is

[07:52] 0.20.

[07:53] Now you're in the neighborhood

[07:56] of 20% of that happening this year.

[08:01] That's the argument for the around 20%

[08:04] case. You basically need your lambda

[08:07] value to be higher. And there are

[08:10] different ways to get there, right? Six

[08:13] bids with a conversion around 3 and a

[08:16] half%. So if you took the original

[08:18] number and you just made your conversion

[08:20] double, well, that's going to get it

[08:22] higher. Maybe eight bids

[08:25] with the same conversion percentage,

[08:26] that'll get it higher as well. So you

[08:29] can experiment with this process. And

[08:32] that's it. That's the process. We didn't

[08:34] start with the vibe. We started with

[08:37] real bid frequencies. These are real

[08:40] numbers, chances and conversion. We're

[08:42] looking at the past decade or so. And

[08:43] then we're adjusting our assumptions

[08:45] based off of how the game is changing.

[08:48] We used a conversion rate that reflects

[08:50] how brutal brutally hard it is to finish

[08:52] these off. Then we stress tested what

[08:55] assump assumptions you need to make to

[08:57] justify something higher. Now you just

[08:59] compare that to the DraftKings line.

[09:01] DraftKings at plus 225 implies about

[09:05] 30.8%.

[09:07] Right? We wrote that down earlier. To

[09:09] justify that number, your lambda value

[09:13] needs to be much higher. I've done the

[09:14] math. It needs to be around.37.

[09:19] That means you need either a lot more

[09:22] chances, serious bids than normal, or

[09:25] you need a much higher conversion rate

[09:28] that jumps way above what history

[09:30] suggests or both or a combination of

[09:32] both. That's why process-wise,

[09:36] plus 225 [snorts]

[09:39] feels aggressive. Not because there's

[09:42] only been 24 ever, but because when you

[09:45] break down the problem into chances and

[09:48] conversion, you can see what has to be

[09:50] true for that number to make sense. Now,

[09:52] could we be wrong? Of course, baseball

[09:55] is weird. We're guessing on a lot of

[09:57] this stuff. That's why this is fun. But

[09:59] this is exactly why I can't stand the

[10:02] the 24 argument because it pretends that

[10:05] there's certainty when there isn't

[10:08] certainty. The better way is define

[10:11] chances, estimate conversion, bracket

[10:13] your assumptions, and then translate

[10:15] that into a probability. Now, I want to

[10:19] hear your number. The exact market for

[10:21] this is now up on our show sponsor,

[10:23] Koshi. Volume still pretty low right

[10:26] now. The market is in its infancy, but

[10:28] I'm interested in tracking it as we get

[10:30] closer to opening day because these

[10:32] markets can move fast when people start

[10:34] paying attention. In the comments, tell

[10:36] me what you think a fair price should be

[10:38] in this market. If you had to hang odds

[10:41] on yes, there will be a perfect game

[10:43] this regular season. What would you make

[10:45] it? And and better better than that,

[10:47] don't just give me a number. Give me

[10:49] what assumptions you would make. Maybe

[10:52] I'm just wrong on the number of chances

[10:54] that'll come up this year or the

[10:55] conversion rate or something like that.

[10:57] How many perfect through six bids do you

[10:59] think we see per year? And what

[11:00] conversion rate do you think is

[11:02] reasonable once one of those bids is

[11:05] alive? If you enjoyed this, hit like,

[11:08] hit that subscribe button, send me the

[11:10] next viral book tweet that you want me

[11:12] to price out as well.

Best Sportsbooks

All Sportsbooks

Recent Stories

Betstamp FAQ's

Best Sportsbooks

All Sportsbooks

Recent Stories

How Online Casinos Are Changing the Way People Place Bets

How a Tennis Sharp Discovered This Wild Edge

Exposing The Worst Betting Advice We Found On Twitter | Presented by Kalshi