Character ComparisonCOMPARE
๐ก Strengths & Features of This Tool
- โจOverwhelming Popularity: The most popular tool on Daphne Checker, loved by frontline players to scrutinize the true strength of their characters.
- โจZero-Noise Pure Comparison: Completely excludes equipment, affinities, and total overcome BP for a flat comparison of base stats only.
- โจExtremely High Reliability: Recalculated using only elite data that has survived strict confidence judgments that do not allow even a margin of error of 1. (You can also change the target data from 'Confidence' in the settings.)
โ Target Selection
โ Settings
Fetched 5 characters from DB (Max 140)
โฆ About Character Comparison โฆ
โ About Extraction Conditions for Other Users' Characters
Because the comparison screen is a tool designed to evaluate 'top-tier individuals' alongside 'your individuals', we apply extremely strict filtering when pulling in data from other users.
Restrictions on Your Characters
- Basically, all matching characters and classes are displayed.
- However, characters set to 'No longer using (Archived)' or those that are 'Empty Shells' will not be displayed unless you check the 'Include' box in the settings on the screen.
Extraction Restrictions on Other Users' Characters (Max 140)
For other users' characters to enter the comparison arena, they must meet all of the following strict conditions:
- They must meet the filters specified on the screen (minimum confidence, fill rate, mental, etc.).
- Predictions for 'all 7 stats' from STR to LUK must be complete (no missing data).
- The predicted value for each of the 7 stats must rank within the 'Top 20'.
* These 'Top 20 for each stat' are combined, duplicates are removed, and only the elite (up to 140 individuals) are summoned from the database for you to compare against.
โ About Extraction Conditions for Other Users' Characters
* The displayed expected values represent pure base stats, with 'Bondmates', 'Shared Discipline BP', and 'Other BP' excluded (subtracted).
* Note that tier bonuses at each level (e.g., Iron Tier Bonus from Lv20, Copper Tier Bonus from Lv40) are added in the calculations and displays.
* The ranking screen sorts by 'Growth Rate', but the values displayed on the comparison screen are evaluated based on actual values including 'Initial BP', so they may not necessarily match the rankings.
โ The Most Used Tool on Our Site
'Character Comparison' is the most popular tool on Daphne Checker, and frontline players frequently use it to scrutinize the true strength of their characters. You might often see screenshots of this comparison screen on social media (X, YouTube, etc.). What you can do here is very simple: 'Compare the stats of characters side-by-side under the exact same conditions.'.
โ The Decisive Difference from 'Typical Stat Screens'
Information like 'Everyone's Level XX Stats' often shared on strategy sites was important for gathering data at the time, but because it mixes in invisible, uncertain elements like 'Bondmates', 'Equipment', 'IVs', and 'Allocated Bonus Points (BP)', it was often unhelpful for pure performance comparisons.
However, this tool is different. The data registered on this site and listed for comparison has been meticulously calculated by reverse-engineering and excluding all the aforementioned uncertain elements. There are strict penalties that immediately drop the confidence rank if there is even the slightest error or contradiction in the input, leaving only the highly reliable data that has survived these checks.
โ Recalculation Under Ultimate Equality
To ensure an equal comparison, 'Base Growth Rates' and 'IVs' are calculated from highly reliable data that has passed strict judgments, and stats are recalculated from scratch based on the desired level set in the comparison table.
In other words, equipment, affinities, or total overcome BP accidentally left on in the game screen are treated as if they 'don't exist at all'. This allows our site to provide the fairest and most accurate character comparisons of any information source.
โ Confidence
The reliability of the predicted results (S to F).
For example, S means an error of 1 BP or less at Level 130, while A means 1 to 3 BP. A lower rank indicates a wider margin of error.
[+][-]Confidence Evaluation Criteria (Details)
Evaluation based on the 'Margin of Error' at Level 130
When the match rate is 100%, the evaluation is based on the 'stat deviation (margin of error) at Lv 130' caused by the variance in the calculated growth rate.
- S : Error of 1 BP or less
- A : Error of 1 to 3 BP
- B : Error of 3 to 5 BP
- C : Error of 5 to 7 BP
- D : Error of 7 to 10 BP
- F : Error of 10 BP or more
* Due to the nature of the calculations, the stats at Level 130 are not finalized until the character actually reaches Level 130.
* As more input data (level-up records) is provided, candidates are narrowed down, and the confidence level will increase towards S.
Match Rate Penalty
The 'Match Rate' with the observed data should inherently be 100%. If it falls below this, it is highly likely there is an error in the data input or prerequisites, resulting in the following penalties:
- Under 100% to 95%: If the error is less than 5, forced demotion to 'D'; if 5 or more, forced demotion to 'F'.
- Under 95%: Forced demotion to 'F' regardless of the error margin.
Penalty for Insufficient Data Input
If a stat has 'less than 5 levels' of input data, and 'there is a tier level bracket (e.g., Unranked Lv 1-20, Lead Lv 21-30) between Level 1 and the highest entered level with zero data points,' the reliability of the prediction significantly drops. Therefore, the confidence of that stat is forcibly downgraded by 2 ranks (e.g., S โ B, B โ D).
* Even if this penalty reduces the confidence to 'F', as a special exception, it will not cause a 'chain-reaction downgrade' to other stats.
Penalty for Abnormal Total BP
If the total BP calculated from the predicted results of all stats is '11 or more' or '4 or less (*when all stats are entered but the total does not reach the minimum of 5)', it is considered an input error or an abnormal value, and the confidence of all stats is forcibly demoted to 'F'.
* This penalty does not apply to special characters, such as the 'Masked Adventurer', who do not have BP allocation mechanics.
Penalty for Statistical Anomalies (Missing Bondmate Settings, etc.)
The system automatically detects 'impossibly high growth rates' possibly caused by forgetting to configure Bondmates in the tool despite having them equipped in-game. The confidence of that stat is forcibly demoted to 'F'.
Statistical upper limits are automatically calculated based on data collected from all users. Data that significantly exceeds this limit (or data that exceeds the maximum limit for a normal individual despite not having IVs) is subject to this penalty.
* Even if this penalty reduces the confidence to 'F', as a special exception, it will not cause a 'chain-reaction downgrade' to other stats.
* If settings are corrected and the values return to normal, the original evaluation will be automatically restored.
* Data where the IV judgment is uncertain (an intermediate state where the probability is not 0% or 100%) is exempt from the penalty.
Chain-Reaction Downgrade from 'F Rank' in Other Stats
For every stat that receives an 'F' confidence rank, the evaluation of all other stats is downgraded by 1 rank.
Example: If the initial evaluation was 'STR(F)', 'INT(F)', and 'Other(S)', since there are two Fs, the 'S' is downgraded by 2 ranks, making the final evaluation 'STR(F)', 'INT(F)', and 'Other(B)'.