Protocol Assimilation Resource Statistics
Disclaimer: numbers presented here are the result of my own data records. They are only an estimation of the actual rates, and individual luck will vary. Statistics here are aggregated from several different accounts to improve accuracy, and so far there does not appear to be any per-account variation in rates outside of what normal luck would explain. Over a sufficiently large period of time, your own results will not be far from these.
Coalition Unit size distribution#
SF unit sizes are not uniformly distributed, and instead follow a sort of bell-shaped curve. Below are the frequencies of each size from my pulls and submitted data:
Interestingly, XL units appear to be more common than XS. The distribution of sizes is not symmetric. In table form, my current best estimates:
| Size | Estimated Rate | Lower bound | Upper bound |
|---|---|---|---|
| Extra Small | 2% | 1.35% | 2.21% |
| Small | 17% | 15.91% | 18.07% |
| Medium | 60% | 58.77% | 61.58% |
| Large | 16% | 15.17% | 17.29% |
| Extra Large | 5% | 4.38% | 5.64% |
Upper and lower bounds are listed with 95% confidence.
So far I see no good evidence to suggest the true rate varies by base rarity (i.e. for 1 or 2 star mooks).
Note: with XL/L mooks having a ~20% total rate and maximum dummy links/analysis taking 5 total copies, I would generally not recommend raising medium or smaller units. On average, if you've pulled enough to fully link it, odds are decent you'll have pulled at least a large. On the other hand, waiting for an XL unit (especially for 2-star mooks) could easily take the better part of a year and may not be worth waiting for. Either way, keep in mind that 99% of coalition development can be done instantly. If you don't have an immediate use for a unit, there's no need to invest in it.
Shiny rate#
"Shiny" units have a golden sparkle on their ID cards, and an animated effect on their frontmost chibi. This is a purely cosmetic effect, and has a small chance of coming with any unit you pull. Toggling the "Cognitive Infusion Button" to "Golden Infusion" in the Symmetric Infusion Drive lets you transfer shiny status to another unit.
I do not currently have enough data to get a firm estimate on the shiny rate. As of last update, I have 25 shinies from 5000 total pulls (0.500% rate). The true rate could plausibly be anywhere from 0.32-0.74% based on this. It is very unlikely (<0.1% chance) for the true shiny rate to be 1%, and 0.5% is a reasonable guess.
One point of caution - the achievements for captures count failed attempts on 2* units and ringleaders, and don't count units acquired through the shop. Calculating your XL/XS/shiny rate based on those achievements will not give the right numbers.
Coalition Drill: Decoding Practice rates#
The chances of getting Basic/Advanced Training Keycodes are also not even. The true rate is commonly claimed to be 25%, but this is not the case. Advanced keycodes are given out between 29.1-32.5% of the time, with 95% confidence. If I had to pick an exact number, I'd currently be most comfortable putting it at 30% (a 2.33:1 ratio).
SF skills take around 1.7:1 Basic:Advanced keycodes to train, so if you fully level all skills you'll build up a surplus of basic keycodes over time. The advanced rate would need to be around 36-38% for the data types to be even. The most efficient svarog bundle only including 1000 basic keycodes doesn't help either.
For more exact numbers:
| Base rarity | Total basic needed | Total advanced needed | Ratio | Percent advanced |
|---|---|---|---|---|
| 1-star | 1000 | 600 | 1.67:1 | 37.5% |
| 2-star | 2000 | 1200 | 1.67:1 | 37.5% |
| 3-star (Ringleader) | 7200 | 4200 | 1.71:1 | 36.8% |
If you were to stop just before the final skill level (SL4/9), it balances somewhat better:
| Base rarity | Total basic needed | Total advanced needed | Ratio | Percent advanced |
|---|---|---|---|---|
| 1-star | 1000 | 200 | 5:1 | 16.7% |
| 2-star | 2000 | 400 | 5:1 | 16.7% |
| 3-star (Ringleader) | 7200 | 2200 | 3.27:1 | 23.4% |
On one hand, you may want to hold off on the final skill level for non-essentials. That would put your advanced keycode spending more in line with its income. On the other, skill leveling is one of the few ways available to improve your coalition echelon's performance - you can't exactly just give them a stronger fairy.
Keep in mind that, over a multi-month period, the range of normal luck will not stray far from these rate estimates. No one has an average long-term advanced keycode rate of 50%, for example. It is virtually impossible to have a surplus of advanced keycodes, unless you spend gems for more in packages, leave SF skills partially leveled, or don't plan to level a significant amount of SF.
Captures & Failures#
Capture probability
2-star coalition units have a stated 50% capture chance, ringleaders 25%.
| Stated rate | Observed rate | Captures | Failures | |
|---|---|---|---|---|
| 2-star | 50% | 50.82% (47-54%) | 404 | 391 |
| 3-star | 25% | 29.41% (21-39%) | 30 | 72 |
It's hard to get a good sample size on ringleader capture attempts, but it seems very likely the stated rates are accurate. Ranges are listed with 95% confidence. Yes, it is entirely possible to fail a coinflip 8 times in a row.
Failure rewards
When you fail to capture a coalition unit, you get some resources as a sort of pity. The potential rewards are:
| 2-star | Ringleader | |||||
|---|---|---|---|---|---|---|
| Item | Tier 1 | Tier 2 | Tier 3 | Tier 1 | Tier 2 | Tier 3 |
| Manpower | 400 | 500 | 800 | 500 | 800 | 1000 |
| Ammo | 400 | 500 | 800 | 500 | 800 | 1000 |
| MRE | 400 | 500 | 800 | 500 | 800 | 1000 |
| Parts | 300 | 400 | 500 | 400 | 500 | 800 |
| Battery | 30 | 40 | 50 | 40 | 60 | 80 |
| SFCR | 80 | 100 | 120 | 100 | 200 | 300 |
| Petri | 80 | 100 | 120 | 100 | 200 | 300 |
| Basic Keycodes | 80 | 100 | 120 | 100 | 200 | 300 |
| Advanced Keycodes | 50 | 80 | 120 | 100 | 150 | 200 |
- The odds of getting each reward tier are not uniform, and seem to have a 25-37.5-37.5% split (tier 1-2-3). Current experimental rates: 93 (26.9%)/122 (35.3%)/131 (37.9%).
- These rewards can be reasonably put into 5 groups, and each group has a ~20% rate. These are:
- Basic resources (MP/Ammo/Rations/Parts); observed 19.9% rate
- Batteries; observed 20.5% rate
- SFCRs; observed 17.1% rate
- Keycodes; observed 20.5% rate (~2.7:1 basic:advanced ratio)
- Petri dishes; observed 22.0% rate
based on 346 samples.
More like this
on popular farming maps
A deep dive