MTG Salvation

Hi! Just wondering if you have an article describing what you mean by some of the terms you use (e.g. hit over miss) and the process you use to get the numbers? Thanks!

I did a full write-up on hit-over-miss when I introduced it, you can read precisely how this figure is calculated Here. Although the set data is of course old, the calculation is exactly the same.

As for my calculations, I can dive into that fairly quickly.
(Edit: I guess it wasn't so quick, but.. now you know!)

All pricing is directly taken from TCGPlayer, using their Market Value USD price. The Market Value price is the best indication for the actual selling point for most cards. The Low-Mid-High system is, and has been, way off in the past because of outliers and targeted manipulation. The Market Value directly looks at what the card has actually been selling at, and is much harder to disrupt.

As for box value, this is a more complex and long calculation that looks at the expected value a box has. I'll run it down using Ravnica Allegiance as our guide.


First, I always calculate cards that are worth less than $0.50 as $0.00. These cards are generally unsellable in the major market and will only overinflate what you might actually expect to get in returns. You should, of course, sell every card you can - but these are minor cards that will often sit unsold for a long time. They can give you profit in the long haul, but should not be noted when looking at more direct returns. This is where I call cards "sellable" or "unsellable" - it specifically means "worth $0.50 or more" or not.

I don't break the values down per-pack because the variability between packs is far too high. You can find it by dividing the box value by the pack count, but this is generally a unneeded and misleading number unless you're somehow buying a large quantity of individual packs. Outside of trying to invest in singles, when you're looking at flipping product and profit margins, boxes are the smallest denominator.

So instead of packs, I note the average Rare Slot value, which has more broad uses. This is an easy calculation of looking at all rares and mythics, and averaging them accordingly. Each individual rare is twice as common as any individual mythic, so we can calculate this by:

(Rare Average)+(Rare Average)+(Mythic Average) / (Rare Count)+(Rare Count)+(Mythic Count)

This gives us a Rare Slot average of an even $2.50.
This is the only number where I don't toss out cards below $0.50.

For calculating the box value, however, we toss out all cards under $0.50, while still dividing by the entire count. Doing this gives us $2.41 per rare slot instead. Multiply this by the 36 pack box size, and we see that the average RNA box has $86.76 in sellable rares and mythics.

This process is repeated for uncommons and commons, which shouldn't be overlooked. They can add up more than you might think, especially if there is a chase uncommons like Fatal Push. Most sets have almost no value in their commons, but outliers happen. Persistent Petitioners is currently sitting at $0.98, is very sellable even for a common, and appears in pretty much every box.

For RNA, sellable uncommons and commons within the box add up to $9.98.

Add all this together, and the basic pack contents within an RNA box averages $96.74.


Lastly comes Foils, which is honestly a giant probability mess. Foils don't follow the same rarity as regular cards, with rare foils appearing at 1 in 6 foils. The actual expected foil average is 1:2:3. In six foils, one will be rare or mythic, two will be uncommon, and three will be common.

Foils also appear at 1-in-6 packs, which is where the nice average of "1 foil rare per box" comes into play (36 packs = 6 foils = 1 foil rare), as well as the "Almost one foil mythic per case" (8 boxes per case, with 1 in 9 rareslot foils being mythic).

This can all be calcuated in a similar fashion to the above. Average the sellable foil commons, then uncommons, and rares. Multiply by their appearance rate, add them together, then divide by the rate at which all foils appear.

Now, the prices between foils varies greatly, and can introduce a large amount of variability and inconsistencies between boxes. Additionally, some $0.50 foils are harder to sell. Although the scarcity of the foil drives the price a bit, it's not always a coveted card. With all of this, I always round down and add in a bit of lower estimation. This is one of the points where I diverge slightly from others who calculate EV. As I always say, time and time again, when variability is high, round everything down to protect yourself.

You could open an RNA box with a foil Shockland, Hydroid, or Domri - cards that inflate the average quite a bit - but chances are you're going to get run of the mill foils that add up to about $10.75.

Add this to our running total, and we arrive at $107.50. This is a few cents short of what's listed above because we've been rounding down to the nearest penny. (Do this a few times, multiply it out to 36 packs, and we're $0.14 off)

It's a long process, but it's just numbers and math. Luckily I know how to automate such things.


But before I stop, we need to step back for a moment, because people often get mislead by the numbers. We've calculated the RNA box "expected value" - the average return of an RNA box, as calculated across a very large number of boxes. Remembering that this is an "average", however, is very important. Boxes differ. One box might be a gold mine, and the next a dud. In order to actually realize the average we calculate, you'll have to buy quite a few boxes.

This is where the hit-over-miss becomes important. If you're looking at buying a box, or even a few, knowing the box consistency is key. A box with a $10 profit margin but a 300% hit-over-miss means that you might average $10 profit in the long run, but individual boxes carry higher risk than normal. Once we approach 400% hit-over-miss, individual boxes seem to be almost as varied as packs. If you're looking into buying boxes, you absolutely have to know your numbers.

OK, so "Hit over Miss" is the calculation of how many dud rares there are vs how many money rares there are... so it kind of indicates "variability of calculated EV" (where lower is better) - That's what has been confusing me - I figured "Hit over Miss" was reflected in the calculation as a percentage in that order (in other words, if Hit over miss is 300%, you are 3 times as likely to get a box that hits vs one that misses - but [as unnatural as it sounds] the percentage actually reflects "misses over hits").

Hit-Over-Miss is calculated exactly as: A "Hit" box's value divided by A "Miss" box's value.
So it, as the name suggests, Hit divided by miss.

A hit box value is defined as:
The expected value of a booster box which contains all of the set's top 3 rares and top 3 mythics.

In RNA, this would be a box that had Hydroid Krasis, Prime Speaker Vannifar, Spawn of Mayhem, Breeding Pool, Blood Crypt, and Stomping Ground. This is a top end box that got all the right cards. Such a box averages $144.41 in value.


A miss is defined as:
The expected value of a booster box which does NOT contain any of the set's top 3 rares and top 3 mythics.

In RNA, this is a box that does not contain any of the above Hit Box's cards. It's the realistic dud we'll see thrown into the mix. Such a box averages $67.71 in value. (This is very high for a Miss value, actually, but RNA is new and inflated)

What it means and why
These two values gives us a range of what's generally realistic of what we'll actually receive. It's not the only way to get a good box or bad box, but it's a very good delimiter for both. Offering just a range value ($77), however, can be misleading or confusing as the position of the top and bottom is very important. Sets can have about the same range value, while offering completely different levels of variation. By dividing the top by the bottom, we can see how the range compares to the total possibilities, and thus a more accurate depiction of variation. Hit-over-miss thus illustrates the consistency in returns better than range. (Standard deviation, etc, is also less easy to correlate directly)

Thus, by dividing 144.41 by 67.71 we get RNA's Hit-Over-Miss of 213%.

This value, alone, doesn't tell us the variation directly. Just like other set values, we tell if a Hit-Over-Miss is high or low mainly by comparing it to the Hit-Over-Miss values of other sets. A set is only "stable" or "unstable" because of how it compares to other sets.

In general: 200% is extremely stable, 300% is decent, and 400% is unstable.

I could just convert this to a "stability score", but I do always try to be concise and clear in what I'm saying and how I got there, even if the method seems unorthodox.

Apologies to the readers who've missed the MMI the past few weeks. Illness and the ongoing GDPR issues have kept me down and busy for a bit. I'm very hopeful to get back on track now and return to a weekly release moving forward, though.