YES(?, 23*c + 12*c^2 + 14)

Initial complexity problem:
1:	T:
		(1, 1)    evalaxstart(a, b, c) -> evalaxentryin(a, b, c)
		(?, 1)    evalaxentryin(a, b, c) -> evalaxbbin(0, b, c)
		(?, 1)    evalaxbbin(a, b, c) -> evalaxbb2in(a, 0, c)
		(?, 1)    evalaxbb2in(a, b, c) -> evalaxbb1in(a, b, c) [ c >= b + 2 ]
		(?, 1)    evalaxbb2in(a, b, c) -> evalaxbb3in(a, b, c) [ b + 1 >= c ]
		(?, 1)    evalaxbb1in(a, b, c) -> evalaxbb2in(a, b + 1, c)
		(?, 1)    evalaxbb3in(a, b, c) -> evalaxbbin(a + 1, b, c) [ b + 1 >= c /\ c >= a + 3 ]
		(?, 1)    evalaxbb3in(a, b, c) -> evalaxreturnin(a, b, c) [ c >= b + 2 ]
		(?, 1)    evalaxbb3in(a, b, c) -> evalaxreturnin(a, b, c) [ a + 2 >= c ]
		(?, 1)    evalaxreturnin(a, b, c) -> evalaxstop(a, b, c)
	start location:	evalaxstart
	leaf cost:	0

Repeatedly removing leaves of the complexity graph in problem 1 produces the following problem:
2:	T:
		(1, 1)    evalaxstart(a, b, c) -> evalaxentryin(a, b, c)
		(?, 1)    evalaxentryin(a, b, c) -> evalaxbbin(0, b, c)
		(?, 1)    evalaxbbin(a, b, c) -> evalaxbb2in(a, 0, c)
		(?, 1)    evalaxbb2in(a, b, c) -> evalaxbb1in(a, b, c) [ c >= b + 2 ]
		(?, 1)    evalaxbb2in(a, b, c) -> evalaxbb3in(a, b, c) [ b + 1 >= c ]
		(?, 1)    evalaxbb1in(a, b, c) -> evalaxbb2in(a, b + 1, c)
		(?, 1)    evalaxbb3in(a, b, c) -> evalaxbbin(a + 1, b, c) [ b + 1 >= c /\ c >= a + 3 ]
	start location:	evalaxstart
	leaf cost:	3

Repeatedly propagating knowledge in problem 2 produces the following problem:
3:	T:
		(1, 1)    evalaxstart(a, b, c) -> evalaxentryin(a, b, c)
		(1, 1)    evalaxentryin(a, b, c) -> evalaxbbin(0, b, c)
		(?, 1)    evalaxbbin(a, b, c) -> evalaxbb2in(a, 0, c)
		(?, 1)    evalaxbb2in(a, b, c) -> evalaxbb1in(a, b, c) [ c >= b + 2 ]
		(?, 1)    evalaxbb2in(a, b, c) -> evalaxbb3in(a, b, c) [ b + 1 >= c ]
		(?, 1)    evalaxbb1in(a, b, c) -> evalaxbb2in(a, b + 1, c)
		(?, 1)    evalaxbb3in(a, b, c) -> evalaxbbin(a + 1, b, c) [ b + 1 >= c /\ c >= a + 3 ]
	start location:	evalaxstart
	leaf cost:	3

A polynomial rank function with
	Pol(evalaxstart) = 3*V_3 - 1
	Pol(evalaxentryin) = 3*V_3 - 1
	Pol(evalaxbbin) = -3*V_1 + 3*V_3 - 1
	Pol(evalaxbb2in) = -3*V_1 + 3*V_3 - 2
	Pol(evalaxbb1in) = -3*V_1 + 3*V_3 - 2
	Pol(evalaxbb3in) = -3*V_1 + 3*V_3 - 3
orients all transitions weakly and the transition
	evalaxbb3in(a, b, c) -> evalaxbbin(a + 1, b, c) [ b + 1 >= c /\ c >= a + 3 ]
strictly and produces the following problem:
4:	T:
		(1, 1)          evalaxstart(a, b, c) -> evalaxentryin(a, b, c)
		(1, 1)          evalaxentryin(a, b, c) -> evalaxbbin(0, b, c)
		(?, 1)          evalaxbbin(a, b, c) -> evalaxbb2in(a, 0, c)
		(?, 1)          evalaxbb2in(a, b, c) -> evalaxbb1in(a, b, c) [ c >= b + 2 ]
		(?, 1)          evalaxbb2in(a, b, c) -> evalaxbb3in(a, b, c) [ b + 1 >= c ]
		(?, 1)          evalaxbb1in(a, b, c) -> evalaxbb2in(a, b + 1, c)
		(3*c + 1, 1)    evalaxbb3in(a, b, c) -> evalaxbbin(a + 1, b, c) [ b + 1 >= c /\ c >= a + 3 ]
	start location:	evalaxstart
	leaf cost:	3

Repeatedly propagating knowledge in problem 4 produces the following problem:
5:	T:
		(1, 1)          evalaxstart(a, b, c) -> evalaxentryin(a, b, c)
		(1, 1)          evalaxentryin(a, b, c) -> evalaxbbin(0, b, c)
		(3*c + 2, 1)    evalaxbbin(a, b, c) -> evalaxbb2in(a, 0, c)
		(?, 1)          evalaxbb2in(a, b, c) -> evalaxbb1in(a, b, c) [ c >= b + 2 ]
		(?, 1)          evalaxbb2in(a, b, c) -> evalaxbb3in(a, b, c) [ b + 1 >= c ]
		(?, 1)          evalaxbb1in(a, b, c) -> evalaxbb2in(a, b + 1, c)
		(3*c + 1, 1)    evalaxbb3in(a, b, c) -> evalaxbbin(a + 1, b, c) [ b + 1 >= c /\ c >= a + 3 ]
	start location:	evalaxstart
	leaf cost:	3

A polynomial rank function with
	Pol(evalaxbb2in) = 1
	Pol(evalaxbb3in) = 0
	Pol(evalaxbb1in) = 1
and size complexities
	S("evalaxbb3in(a, b, c) -> evalaxbbin(a + 1, b, c) [ b + 1 >= c /\\ c >= a + 3 ]", 0-0) = 3*c + 9
	S("evalaxbb3in(a, b, c) -> evalaxbbin(a + 1, b, c) [ b + 1 >= c /\\ c >= a + 3 ]", 0-1) = ?
	S("evalaxbb3in(a, b, c) -> evalaxbbin(a + 1, b, c) [ b + 1 >= c /\\ c >= a + 3 ]", 0-2) = c
	S("evalaxbb1in(a, b, c) -> evalaxbb2in(a, b + 1, c)", 0-0) = 3*c + 9
	S("evalaxbb1in(a, b, c) -> evalaxbb2in(a, b + 1, c)", 0-1) = ?
	S("evalaxbb1in(a, b, c) -> evalaxbb2in(a, b + 1, c)", 0-2) = c
	S("evalaxbb2in(a, b, c) -> evalaxbb3in(a, b, c) [ b + 1 >= c ]", 0-0) = 3*c + 9
	S("evalaxbb2in(a, b, c) -> evalaxbb3in(a, b, c) [ b + 1 >= c ]", 0-1) = ?
	S("evalaxbb2in(a, b, c) -> evalaxbb3in(a, b, c) [ b + 1 >= c ]", 0-2) = c
	S("evalaxbb2in(a, b, c) -> evalaxbb1in(a, b, c) [ c >= b + 2 ]", 0-0) = 3*c + 9
	S("evalaxbb2in(a, b, c) -> evalaxbb1in(a, b, c) [ c >= b + 2 ]", 0-1) = ?
	S("evalaxbb2in(a, b, c) -> evalaxbb1in(a, b, c) [ c >= b + 2 ]", 0-2) = c
	S("evalaxbbin(a, b, c) -> evalaxbb2in(a, 0, c)", 0-0) = 3*c + 9
	S("evalaxbbin(a, b, c) -> evalaxbb2in(a, 0, c)", 0-1) = 0
	S("evalaxbbin(a, b, c) -> evalaxbb2in(a, 0, c)", 0-2) = c
	S("evalaxentryin(a, b, c) -> evalaxbbin(0, b, c)", 0-0) = 0
	S("evalaxentryin(a, b, c) -> evalaxbbin(0, b, c)", 0-1) = b
	S("evalaxentryin(a, b, c) -> evalaxbbin(0, b, c)", 0-2) = c
	S("evalaxstart(a, b, c) -> evalaxentryin(a, b, c)", 0-0) = a
	S("evalaxstart(a, b, c) -> evalaxentryin(a, b, c)", 0-1) = b
	S("evalaxstart(a, b, c) -> evalaxentryin(a, b, c)", 0-2) = c
orients the transitions
	evalaxbb2in(a, b, c) -> evalaxbb3in(a, b, c) [ b + 1 >= c ]
	evalaxbb2in(a, b, c) -> evalaxbb1in(a, b, c) [ c >= b + 2 ]
	evalaxbb1in(a, b, c) -> evalaxbb2in(a, b + 1, c)
weakly and the transition
	evalaxbb2in(a, b, c) -> evalaxbb3in(a, b, c) [ b + 1 >= c ]
strictly and produces the following problem:
6:	T:
		(1, 1)          evalaxstart(a, b, c) -> evalaxentryin(a, b, c)
		(1, 1)          evalaxentryin(a, b, c) -> evalaxbbin(0, b, c)
		(3*c + 2, 1)    evalaxbbin(a, b, c) -> evalaxbb2in(a, 0, c)
		(?, 1)          evalaxbb2in(a, b, c) -> evalaxbb1in(a, b, c) [ c >= b + 2 ]
		(3*c + 2, 1)    evalaxbb2in(a, b, c) -> evalaxbb3in(a, b, c) [ b + 1 >= c ]
		(?, 1)          evalaxbb1in(a, b, c) -> evalaxbb2in(a, b + 1, c)
		(3*c + 1, 1)    evalaxbb3in(a, b, c) -> evalaxbbin(a + 1, b, c) [ b + 1 >= c /\ c >= a + 3 ]
	start location:	evalaxstart
	leaf cost:	3

A polynomial rank function with
	Pol(evalaxbb2in) = -2*V_2 + 2*V_3 + 1
	Pol(evalaxbb1in) = -2*V_2 + 2*V_3
and size complexities
	S("evalaxbb3in(a, b, c) -> evalaxbbin(a + 1, b, c) [ b + 1 >= c /\\ c >= a + 3 ]", 0-0) = 3*c + 9
	S("evalaxbb3in(a, b, c) -> evalaxbbin(a + 1, b, c) [ b + 1 >= c /\\ c >= a + 3 ]", 0-1) = ?
	S("evalaxbb3in(a, b, c) -> evalaxbbin(a + 1, b, c) [ b + 1 >= c /\\ c >= a + 3 ]", 0-2) = c
	S("evalaxbb1in(a, b, c) -> evalaxbb2in(a, b + 1, c)", 0-0) = 3*c + 9
	S("evalaxbb1in(a, b, c) -> evalaxbb2in(a, b + 1, c)", 0-1) = ?
	S("evalaxbb1in(a, b, c) -> evalaxbb2in(a, b + 1, c)", 0-2) = c
	S("evalaxbb2in(a, b, c) -> evalaxbb3in(a, b, c) [ b + 1 >= c ]", 0-0) = 3*c + 9
	S("evalaxbb2in(a, b, c) -> evalaxbb3in(a, b, c) [ b + 1 >= c ]", 0-1) = ?
	S("evalaxbb2in(a, b, c) -> evalaxbb3in(a, b, c) [ b + 1 >= c ]", 0-2) = c
	S("evalaxbb2in(a, b, c) -> evalaxbb1in(a, b, c) [ c >= b + 2 ]", 0-0) = 3*c + 9
	S("evalaxbb2in(a, b, c) -> evalaxbb1in(a, b, c) [ c >= b + 2 ]", 0-1) = ?
	S("evalaxbb2in(a, b, c) -> evalaxbb1in(a, b, c) [ c >= b + 2 ]", 0-2) = c
	S("evalaxbbin(a, b, c) -> evalaxbb2in(a, 0, c)", 0-0) = 3*c + 9
	S("evalaxbbin(a, b, c) -> evalaxbb2in(a, 0, c)", 0-1) = 0
	S("evalaxbbin(a, b, c) -> evalaxbb2in(a, 0, c)", 0-2) = c
	S("evalaxentryin(a, b, c) -> evalaxbbin(0, b, c)", 0-0) = 0
	S("evalaxentryin(a, b, c) -> evalaxbbin(0, b, c)", 0-1) = b
	S("evalaxentryin(a, b, c) -> evalaxbbin(0, b, c)", 0-2) = c
	S("evalaxstart(a, b, c) -> evalaxentryin(a, b, c)", 0-0) = a
	S("evalaxstart(a, b, c) -> evalaxentryin(a, b, c)", 0-1) = b
	S("evalaxstart(a, b, c) -> evalaxentryin(a, b, c)", 0-2) = c
orients the transitions
	evalaxbb2in(a, b, c) -> evalaxbb1in(a, b, c) [ c >= b + 2 ]
	evalaxbb1in(a, b, c) -> evalaxbb2in(a, b + 1, c)
weakly and the transition
	evalaxbb2in(a, b, c) -> evalaxbb1in(a, b, c) [ c >= b + 2 ]
strictly and produces the following problem:
7:	T:
		(1, 1)                  evalaxstart(a, b, c) -> evalaxentryin(a, b, c)
		(1, 1)                  evalaxentryin(a, b, c) -> evalaxbbin(0, b, c)
		(3*c + 2, 1)            evalaxbbin(a, b, c) -> evalaxbb2in(a, 0, c)
		(6*c^2 + 7*c + 2, 1)    evalaxbb2in(a, b, c) -> evalaxbb1in(a, b, c) [ c >= b + 2 ]
		(3*c + 2, 1)            evalaxbb2in(a, b, c) -> evalaxbb3in(a, b, c) [ b + 1 >= c ]
		(?, 1)                  evalaxbb1in(a, b, c) -> evalaxbb2in(a, b + 1, c)
		(3*c + 1, 1)            evalaxbb3in(a, b, c) -> evalaxbbin(a + 1, b, c) [ b + 1 >= c /\ c >= a + 3 ]
	start location:	evalaxstart
	leaf cost:	3

Repeatedly propagating knowledge in problem 7 produces the following problem:
8:	T:
		(1, 1)                  evalaxstart(a, b, c) -> evalaxentryin(a, b, c)
		(1, 1)                  evalaxentryin(a, b, c) -> evalaxbbin(0, b, c)
		(3*c + 2, 1)            evalaxbbin(a, b, c) -> evalaxbb2in(a, 0, c)
		(6*c^2 + 7*c + 2, 1)    evalaxbb2in(a, b, c) -> evalaxbb1in(a, b, c) [ c >= b + 2 ]
		(3*c + 2, 1)            evalaxbb2in(a, b, c) -> evalaxbb3in(a, b, c) [ b + 1 >= c ]
		(6*c^2 + 7*c + 2, 1)    evalaxbb1in(a, b, c) -> evalaxbb2in(a, b + 1, c)
		(3*c + 1, 1)            evalaxbb3in(a, b, c) -> evalaxbbin(a + 1, b, c) [ b + 1 >= c /\ c >= a + 3 ]
	start location:	evalaxstart
	leaf cost:	3

Complexity upper bound 23*c + 12*c^2 + 14

Time: 0.189 sec (SMT: 0.177 sec)