maybe this tcrs script actually produced a real result?

int[int] m;
file_to_map("mt_rand 1 to 170 first 1000000 seeds.txt",m);
int c;
boolean b;
void butt(int d, int[int] x ,int r){
	//clear(m);
	print("seeds where mt_rand 1 to 170 is "+r+":");
	for i from 1 to (500000){
		if((m[i]) == r){
			c = 0;
			b = true;
			for j from 0 to (count(x)-1){
				b = b && (m[i+c] == m[i+c+x[j]]);
				c += x[j];
			}
			if(b){
				print(i+" = "+d+" * " + (i/d) + " + " + (i%d) );
				//abort("we did it");
			}
		}
	}
	print("====================");


	/*
	print("seeds where glibc rand 0 to 2147483647 is "+r+":");
	file_to_map("glibc rand 0 to 2147483647 first 1000000 seeds mod 170 plus 1.txt",m);
	for i from 1 to (500000){
		if((m[i]) == r){
			c = 0;
			b = true;
			for j from 0 to (count(x)-1){
				b = b && (m[i+c] == m[i+c+x[j]]);
				c += x[j];
			}
			if(b){
				print(i+" = "+d+" * " + (i/d) + " + " + (i%d) );
				abort("we did it");
			}
		}
	}
	print("====================");
	*/

	/*
	print("seeds where mt_rand mod 170 is "+r+":");
	clear(m);
	file_to_map("mt_rand mod 170.txt",m);
	for i from 1 to (500000){
		if((m[i]+1) == r){
			c = 0;
			b = true;
			for j from 0 to (count(x)-1){
				b = b && (m[i+c] == m[i+c+x[j]]);
				c += x[j];
			}
			if(b){
				print(i+" = "+d+" * " + (i/d) + " + " + (i%d) );
				abort("we did it.");
			}
		}
	}
	print("====================");
	*/
}
//152 = 19*2^3
//245 = 5*7*7
int[int] but = {152*3,245*3,451*3};
/*these have precisely the first enchantment for class 1 path 1.
532	804245262
777	697647321
1331	718230343
3066	401867098
4667	933430014
6959	195097050
8180	121809700

with differences:
245
554
1735
1601
2292
1221
*/

/*these have precisely the second enchantment for class 1 path 1.

94	606216746
1275	121934420
2839	122462796
3519	767266958
4191	851515177
4472	218609014
6041	452627138
9839	880434260

with differences:
1181
1564
680
672
281
1569
3798
*/
/* items with precisely first modifier for class 2 sign 1
191
1779
3508
4442
7541
9576
*/

/* items with precisely first modifier for class 3 sign 1

696
1933
2835
2959
3314
4180
4265
*/

/* items with precisely first modifier for class 1 sign 2

3093
4751
4958
5956
6929
7752
*/

/* items with precisely first modifier for class 3 sign 4

149
2892
4590
4681

*/

for ii from 50 to 50{
	//but = {152*ii,245*ii,451*ii,103*ii};
	//but = {245*ii,554*ii,1735*ii,1601*ii,2292*ii,1221*ii};
	//but = {1181*ii,1564*ii,680*ii,672*ii,281*ii,1569*ii,3798*ii};
	//but = {245*ii,554*ii,1735*ii,1601*ii,2292*ii,1221*ii};
	//but = {(1779-191)*ii,(3508-1779)*ii,(4442-3508)*ii,(7541-4442)*ii,(9576-7541)*ii};
	//but = {(1933-696)*ii,(2835-1933)*ii,(2959-2835)*ii,(3314-2959)*ii,(4180-3314)*ii,(4265-4180)*ii};
	//but = {(4751-3093)*ii,(4958-4751)*ii,(5956-4958)*ii,(6929-5956)*ii,(7752-6929)*ii};
	but = {(2892-149)*ii,(4590-2892)*ii,(4681-4590)*ii};
	print(ii);
	for i from 1 to 170{
		//butt(532,but,i);
		//butt(3093,but,i);
		butt(149,but,i);
	}
}





//success for but = {1520,2450,4510}:
//seeds where glibc rand 0 to 2147483647 is 3:
//13841 = 380 * 36 + 161
//thus seeds for items 380, 532, 777, and 1228 are 13841, 15361, 17811, and 22321 respectively for seal clubber under mongoose?
//this is likely a false positive; the pattern doesn't hold for the next item in the sequence.
//I think the reason is that some of the items had multiple modifiers. we'll exclude those in future tests.



//success for but = {12250,27700,86750,80050,114600,61050}:
//seeds where mt_rand 1 to 170 is 168:
//138955 = 532 * 261 + 103
/*thus seeds for items
532
777
1331
3066
4667
6959
8180

are

138955
151205
178905
265655
345705
460305
521355

respectively.
*/
//horrifyingly, that suggests 18 + 50*itemID + 112337 as the argument for mt_srand().

//but = {1181*ii,1564*ii,680*ii,672*ii,281*ii,1569*ii,3798*ii} produces this, too:
//seeds where mt_rand 1 to 170 is 167:
//117055 = 94 * 1245 + 25
//which has an identical seeding function, thus we really did it.