To Fee or not to Fee

[organofcorti]

This might help. I've excluded all the txs that were included in the next block, since most of them were.



         n  meanTime medianTime feePerByte
 [1,] 40969 245.57797      221.0          0
 [2,] 10445 226.43025      195.0          1
 [3,]  1334 191.66117      194.0          2
 [4,]   261 118.05747       75.0          3
 [5,]   277 130.15162       96.0          4
 [6,]    73 113.31507       46.0          5
 [7,]    46 145.69565       48.5          6
 [8,]    49  49.73469       39.0          7
 [9,]    56  97.51786       25.0          8
[10,]    15  83.13333       39.0          9
[11,]    22  55.68182       39.0         10
[12,]    35  94.05714       45.0         11
[13,]     7  60.71429       38.0         12
[14,]    18 100.72222       23.0         13
[15,]     6  41.66667       33.0         14
[16,]     6  54.00000       24.0         15
[17,]     4  10.50000       10.0         16
[18,]     1 148.00000      148.0         17
[19,]     1  11.00000       11.0         21
[20,]     1   7.00000        7.0         22
[21,]     2 140.50000      140.5         24
[22,]     6  32.83333       45.0         34

[1714131078]

1714131078

[deepceleron]

The analysis to do is on no-fee transactions that have a priority > 57.6M. Compare with the times of those < 57.6M which include exactly the minimum per int(KB+1) fee. These are the default sending modes of Bitcoin-qt with no optional fee.

My hypothesis is that Bitcoin's default behaviour creates two classes of transactions, where the "preferred/non-spammy" class actually gets a lower quality of service from miners.