Pokemon Bracket Fusion Project – Fusion Generation 1 Stats Round-up
Welcome to the first PBFP stats round-up! I thought it’d be fun to see a break down of how the bracket progresses and take some guesses into how fusions or the final fusion might look!
You can find more information on this project and a list of completed fusions here!
Best/Worst Matches
Starting off with the best (and worst!) pairings in the project!
Best Overall Pair
To no surprise, Plusle and Minun take this trophy home by being best matches with a score of 100%! This score is impossible to repeat in future fusion generations given the wide variety of fusion results from the first pairing.
Worst Overall Pair
There’s two different ways to weight “worst” overall pair. The worst score is Burmy and Flabebe at 46%, but the worst match is Hoopa and Xurtitree with each other being #106 and #298 (out of 421) on each other’s lists respectively! Two unfortunate pairs for sure!
Best Match Stats
91/210 (43.33%) Fusions were Best Matches – The number of fusions where both parent Pokemon were paired with their first choice. These Pokemon had sights for no others, just each other <3
44/210 (20.95%) Fusions were Partial Best Matches – The number of fusions where only one parent was paired with their Best Match. The worst partial best match is Arceus and Furfrou. Arceus’s first choice was Furfrou, but Furfrou went through 12 other candidates before settling for Arceus, ouch.
Bye
Because there were an odd number of families, one was unpaired. I assumed one of the more obviously outliers would get the honor (like Flabebe’s forms, Arceus’s forms, or Eevee’s evolutions) but no, it was Frillish. Rather than being ignored, Frillish is added to the fusion gen 2 pool and gets first dibs on a partner.
Tags
Certain families got tags if they fell under specified categories, with the aim of seeing which resulting fusions still fall under those categories. These tags had no weight on how fusions were paired, it just turns out that Pokemon with similar tags could be more likely to pair based on their shared attributes. The tags I’m using are: Starter, Fossil, Pseudo-Legendary, Legendary, Mythical and Ultra Beast.
I used the following rules to determined how the resulting fusions get these tags:
- If both parents have the tag, the fusion gets the tag (so two Fossil parents = a Fossil fusion).
- Any combination of two Legendary Pokemon, Mythical Pokemon, or Ultra Beasts (that are not matching under the first rule) results in a legendary fusion.
- A tagged Pokemon + an untagged Pokemon = untagged Pokemon.
- (under consideration for fusion generation 2 and beyond) If a tagged fusion pairs with a formerly tagged fusion, the resulting fusion gets the tag back (ex. a Fossil/Fossil fusion pairs with a Fossil/non-Fossil fusion, then the resulting fusion is a Fossil fusion).
Because of these rules, it’s near impossible for a tag to exist on the final fusion, so I’m more curious to see how long they last. Speaking of, the tag breakdown so far!
The Ultra Beasts are completely wiped out. This isn’t surprising, as given their diversity they were unlikely to pair up with one another.
Mythicals and Pseudo-Legendaries are likely to follow next gen, as there’s only two left: Vicchi and Phiphy; Myni and Deigon.
Starters seem to be doing better at first glance (there’s 6 left) but there’s actually only 2 of each type (Grass: Chikotwig, Pincko; Fire: Torchar, Chimpig; Water: Wottdile, Pipkie).
Fossils have a bit of a better chance, as there’s 3 left: Tounyte, Lilaura, and Artyl.
Legendaries have the best chance of sticking around, having 15 members remaining. This is largely in part due to rule 2, where Legendary is being treated as a higher level category to Mythical or Ultra Beast.
Stages/Forms
I determine stages and forms by averaging the parents’ values. So fusions of Pokemon families with the same number of stages/forms use those, but fusions of Pokemon families with a different number of stages/forms use the average of the family’s stats, rounded towards the overall average. The overall average is from the whole population for each fusion generation. For the first fusion generation, the overall average for stages was 1.92 and 2.51 for forms (so a fusion of a 1 stage and 2 stage line would be 1.5 stages, then round up to 2 stages because it is less than the overall average of 1.92).
Stages
To no surprise, Pokemon’s stages will move towards the average and away from the extremes. My guess at this time is the final fusion will be a two-stage line.
Forms
Let’s start with the graph since it looks similar to the stage graph.
As the overall average is 2.51, we see fusions with 2 or 3 forms increasing in number. But something to note here is this count is specifically including “base” forms. So a Pokemon with only one stage and zero alternate forms would appear as having 1 (base) form here. This is just related to how I paired things, to where Pokemon with the exact same layout of stages and forms would be the best to pair, and second best would be those very close in stages/forms. Ignoring base forms in the calculations resulted in odd pairings.
Additionally, the overall number of Pokemon with forms is on the rise! Original Pokemon (up to Gen 7) had 22.8% of Pokemon families with forms. The first fusion generation ended up with 25.71% of fusion families with forms!
The overall averages also changed from 1.92 stages to 1.93 stages and 2.51 forms to 2.53 forms. Not too much.
Type
Similarly to stages/forms, I averaged the parents’ types to determine the fusion’s types. Parents with similar types result in a fusion with those same types. For unmatching types:
- Water + Water/Flying = Water/Flying
- Water + Fire = Water/Fire
- Water/Flying + Fire = Fire/???
- Water/Flying + Fire/Flying = Flying/???
- Water/Flying + Fire/Electric = ???/???
For conflicts (the ??? above) I subjectively pick the parent’s type I feel best represents the fusion. If the fusion has multiple stages/forms, I’ll sometimes try to average out the types (so for a Water/Flying + Fire/Flying fusion with 2 forms, I may try to have one form be Water/Flying and the other be Fire/Flying). I’ve internally referred to this as the Zamolt problem, as Zapdos/Moltres’s fusion was the first case of this (Electric/Flying + Fire/Flying fusion). I choose Flying/Fire for the final combination, as being on fire is a stronger indicator of type than being yellow and spiky.
For hypothesizing, we’re likely to see the most common types remain while the least common are slowly killed off. But because of the above, strongly “typed” types (like fire = fire type) are more likely to hang on when in competition.
Statistically, we can see that happening to a small degree. Normal (one of the plainest types) dropped from 9.85% to 9.59%. It wasn’t the largest drop though, as Poison dropped from 4.31% to 3.76%. Psychic had the biggest increase, 9.47% to 10.24%, overtaking Water as the most common type. This may have to do with the large number of pure Psychic types, but that doesn’t explain Water’s drop from 10.28% to 9.96%. Ice was, and still is, the least populated type, yet its percentage increased from 2.89% to 3.29%. Curious to see if it keeps up that upwards trend or loses steam.
If I had to take a guess now, I would guess the final fusion will be Psychic/Water typed. Though, as it is likely to have at least one stage or form which could have a different type combination, a third type replacing either Psychic or Water on one of the stages/forms seems likely.
Colors
Colors are interesting for two reasons:
- Because hues can loop, it’s unlikely until the very end for a hue to truly die off. If no green Pokemon exists, but a blue and a yellow Pokemon pair, then green is back.
- But also for condensing the colors into a visual graph, I used the PokeDex colors which are bogus.
For the (less interesting sounding) first point, I determine the fusion’s colors by doing an altered average of the HSL (hue, saturation, lightness) values of the parent’s colors. Using HSL instead of RGB values, I avoid everything quickly turning into a muddy brown or grey and keep things bright and colorful (my color scheme preference and more aligning with Pokemon’s color choices). It outputs some interesting colors on occasion, but overall it’s fine.
While the hue cannot officially die off until there’s too few fusions at the end, the law of averages will still apply to saturation and lightness and we should expect these values to move towards the center (a semi-saturated, medium shade). We can see this with black and white, which dropped from 6.82% to 5.08% and 10.70% to 7.33%, respectively. I expected grey to follow this trend, but the PokeDex’s color assignment has a habit of assigning some more saturated colors to grey than others might (which I tried to mimic) and if a black Pokemon and a white Pokemon fused, a grey fusion would result, so that actually grew from 9.85% to 11.47%.
Speaking of dex colors, I hate ‘em.
Mapping the variety of colors from the resulting fusions to the PokeDex colors was a chore and I am filled with regret. Sure, at first glance it makes sense: white, grey, and black cover the desaturated colors; red, yellow, green, blue, and purple cover the hues. There’s quite a few pink and brown Pokemon, so those get their own entries.
But what about orange?
Did you ever notice orange isn’t a PokeDex color?
Have you ever stared at the most orange Pokemon in your life and had a crisis about whether it should be classified a red, yellow, or brown? Repeatedly!?
You haven’t?
I envy you…
And with that unceremonious ending, that’s the first fusion generation stats! Thank you so much for following along for the first half!
I’ve got some other projects I put on hold to wrap this up, but once those are done I’ll be back with the next generation of fusions! Who will be paired up with who? You’ll just have to wait and see!
[…] 1, but wanted didn’t want to keep redoing everything so we’re just moving forward lol. There’s a further breakdown here, but TLDR I determine the number of stages and forms by taking the average of the parents, and […]