Good - now I need to search for "Thraben Charm" specifically, and also search for monarch/initiative cards. Let me also get the archetype entries and build the main analysis query.
Good, 447 entries. Now let me search for the monarch/initiative cards that appear in CawGate, and build the main card-count analysis queries. I'll run the key card queries in parallel.
The "--" comments are being blocked. Let me rewrite queries without inline comments.
Good. Now let me run all the remaining card queries in parallel:
Now let me get the "Remove Soul" sideboard query and check for initiative variants in CawGate, and also get sources:
I now have all the data I need. Let me compile everything into a comprehensive report. Note that Islands and Plains showed no sideboard usage (all 447 entries had 0 SB), and Fire // Ice and Thraben Charm have very small samples. Palace Sentinels is the Monarch card (initiative cards didn't show up). Let me also check on the Thraben Inspector vs Thraben Charm distinction the user requested.
Based on 447 CawGate (Caw Gates) entries across 190 Pauper tournaments (54% Melee, 45% MTGO, 0.5% Cardsrealm), here is the full breakdown. Winrates exclude draws. "n=" is number of entries at that count. Omitted counts had zero entries.
Overall archetype baseline: 52.4% (979W-888L-178D)
PREORDAIN
MD:
0 copies: 51.9% (n=334, 748W-694L)
1 copy: 62.5% (n=9, 15W-9L)
2 copies: 59.0% (n=47, 118W-82L)
3 copies: 41.9% (n=23, 31W-43L)
4 copies: 52.8% (n=34, 67W-60L)
SB: No entries played Preordain in the sideboard.
RELIC OF PROGENITUS
MD: Nearly all entries played 0 (443, 52.4%); 4 copies: 64.3% (n=3, tiny sample); 2 copies: 0.0% (n=1).
SB:
0 copies: 50.2% (n=298, 620W-615L)
1 copy: 59.1% (n=14, 39W-27L)
2 copies: 55.8% (n=99, 223W-177L)
3 copies: 58.7% (n=30, 81W-57L)
4 copies: 57.1% (n=6, 16W-12L)
DISPEL
MD:
0 copies: 53.9% (n=228, 536W-458L)
1 copy: 52.2% (n=104, 200W-183L)
2 copies: 50.0% (n=101, 217W-217L)
3 copies: 46.4% (n=14, 26W-30L)
SB:
0 copies: 52.5% (n=319, 714W-645L)
1 copy: 52.5% (n=62, 124W-112L)
2 copies: 54.1% (n=54, 125W-106L)
3 copies: 40.0% (n=8, 14W-21L)
4 copies: 33.3% (n=4, 2W-4L)
ENVELOP
MD:
0 copies: 52.5% (n=434, 952W-862L)
1 copy: 41.7% (n=5, 5W-7L)
2 copies: 53.7% (n=8, 22W-19L)
SB:
0 copies: 52.9% (n=366, 793W-705L)
1 copy: 51.8% (n=19, 44W-41L)
2 copies: 50.4% (n=53, 124W-122L)
3 copies: 47.4% (n=9, 18W-20L)
NEGATE
MD: Almost never played MD (446 entries at 0, 52.4%); 1 copy, n=1: 50.0%.
SB:
0 copies: 52.8% (n=433, 962W-860L)
1 copy: 36.7% (n=10, 11W-19L)
2 copies: 40.0% (n=4, 6W-9L)
SPELL PIERCE
MD:
0 copies: 52.0% (n=302, 644W-594L)
1 copy: 56.0% (n=65, 158W-124L)
2 copies: 52.8% (n=68, 158W-141L)
3 copies: 37.2% (n=11, 16W-27L)
4 copies: 60.0% (n=1, trivial sample)
SB:
0 copies: 52.4% (n=441, 970W-880L)
1-3 copies: very small samples, no clear signal.
REMOVE SOUL
MD: Almost never played (446 entries at 0, 52.4%); 1 copy n=1: 66.7%.
SB:
0 copies: 52.1% (n=432, 938W-864L)
1 copy: 57.1% (n=6, 16W-12L)
2 copies: 67.6% (n=8, 25W-12L)
4 copies: n=1 (0W-0L, no data)
PALACE SENTINELS (Monarch variant; initiative cards found zero entries)
MD: Only 2 entries ran it MD (50.0% sample too small).
SB:
0 copies: 51.9% (n=409, 879W-815L)
1 copy: 50.0% (n=18, 33W-33L)
2 copies: 62.6% (n=20, 67W-40L)
ISLAND
MD:
0 copies: 52.7% (n=17, 39W-35L)
1 copy: 64.7% (n=12, 33W-18L)
2 copies: 55.8% (n=14, 29W-23L)
3 copies: 49.4% (n=121, 254W-260L)
4 copies: 53.1% (n=282, 624W-552L)
SB: No entries ran Islands in the sideboard.
PLAINS
MD:
0 copies: 52.9% (n=357, 799W-710L)
1 copy: 51.3% (n=76, 155W-147L)
2 copies: 43.5% (n=7, 10W-13L)
3+ copies: very small samples, no signal.
SB: No entries ran Plains in the sideboard.
DEEP ANALYSIS
MD:
0 copies: 52.7% (n=314, 685W-616L)
1 copy: 51.3% (n=117, 252W-239L)
2 copies: 57.1% (n=15, 40W-30L)
3 copies: 40.0% (n=1, trivial)
SB: Only 1 entry, trivial sample.
THRABEN CHARM
MD:
0 copies: 50.9% (n=49, 89W-86L)
1 copy: 56.7% (n=9, 17W-13L)
2 copies: 52.4% (n=248, 551W-501L)
3 copies: 53.4% (n=134, 306W-267L)
4 copies: 43.2% (n=7, 16W-21L)
SB:
0 copies: 52.6% (n=352, 792W-715L)
1 copy: 53.9% (n=74, 158W-135L)
2 copies: 46.6% (n=17, 27W-31L)
3 copies: 33.3% (n=3, 2W-4L)
4 copies: 0.0% (n=1, 0W-3L)
FIRE // ICE
MD:
0 copies: 51.9% (n=432, 935W-865L)
1 copy: 71.4% (n=1, trivial)
2 copies: 65.0% (n=14, 39W-21L)
SB: Zero entries ran Fire // Ice in the sideboard.
KEY TAKEAWAYS
Relic of Progenitus SB is the clearest positive signal: any copies (1-4) outperform the 0-copy baseline by 5-8 points, with the 0-copy group sitting at only 50.2%. This is among the most data-rich findings (149 entries with Relic SB).
Remove Soul SB at 2 copies shows 67.6% (n=8) and is the biggest positive outlier, though the sample is small.
Palace Sentinels SB at exactly 2 copies shows a notable 62.6% (n=20) versus 51.9% for 0 copies. 1 copy doesn't move the needle at 50.0%.
Fire // Ice MD at 2 copies shows 65.0% (n=14), a meaningful positive signal for a 14-entry sample.
Preordain MD shows a cliff: 2 copies peaks at 59.0% before dropping to 41.9% at 3 copies, suggesting 2 is the ceiling.
Dispel MD shows a linear degradation: 53.9% at 0, 50.0% at 2, 46.4% at 3. More copies correlates with worse outcomes.
Spell Pierce MD: 1-2 copies (56% / 52.8%) better than 0 (52.0%); 3 copies collapses to 37.2%.
Island count: 4 Islands (n=282, the majority) shows 53.1%, while the very small groups running 1-2 show higher rates but are not representative.
Note: Thraben Charm here refers to the white instant from the database. The user may have meant Thraben Inspector instead (a common CawGate creature); if so, please clarify and I can pull that data separately.
Sources: Data drawn from 190 Pauper tournaments featuring Caw Gates entries, split approximately 54% Melee, 45% MTGO, 0.5% Cardsrealm.
https://www.mtgo.com/decklist/pauper-challenge-32-2026-02-2812834072
https://www.mtgo.com/decklist/pauper-challenge-32-2026-02-2712834057
https://melee.gg/Decklist/View/14a34ef7-f1b7-49c5-94f4-b3ff00e0b068