Most rockets have a few failures early on. Any rocket can be made fairly reliable with enough testing. My main concern with the use of large SRBs is their relatively high processing cost. They are both hazardous and heavy from the moment they arrive. A dropped SRB segment would probably ignite, destroying the VAB. The Ares will require maintaining the entire VAB and SRB processing infrastructure, and the impressive but expensive MLPs and crawlers. In contrast the EELVs can be transported at the pad on rubber tired vehicles and/or rails at much lower cost. There are many tedious crane lifts and many hours of fairly hazardous labor in assembling the fueled segments, including such things as hand-positioning the O-rings under a segment suspended on a crane, which I believe required an OSHA waiver. So far as I know the question of reusability has not even been decided for the Ares SRB; certainly reuse is expensive for the Shuttle. For the Shuttle, the SRBs saved cost in development, but added cost to each flight. The largest SRB other than the Shuttle is on the Ariane, and it is far smaller. I simply don't see any prospect for the Ares bringing the cost of human spaceflight down to a practical level. <br /><br />Personally I feel the only cost-effective solids are the "small" SRBs used on the Delta and Atlas which use monolithic cases and nonsteerable nozzles, and can be shipped and attached in one piece; attaching various numbers or SRBs allows some flexibility in payload requirements. Even then, the all-liquid Delta IV-H, despite its teething problems, has higher performance. <br />